Practical perspectives on agentic AI, systems and evidence.https://softlander.co/en-gbSat, 26 Sep 2026 17:37:03 GMThttps://softlander.co/insights/blog/identity-and-tool-permission-boundaries/https://softlander.co/insights/blog/identity-and-tool-permission-boundaries/An agent can only be evaluated in the context of what its tools allow. Examine identity, delegated authority and action boundaries before you put a workflow in motion.Sat, 12 Sep 2026 10:00:00 GMTGaurav TiwariAgent securityArchitecture & engineeringAgent authorityTool securityDecisions behind dependable agentshttps://softlander.co/insights/blog/defining-agent-evaluation-evidence/https://softlander.co/insights/blog/defining-agent-evaluation-evidence/A release decision needs a traceable connection between expected behaviour, test conditions and observed results. Here is a practical way to build that connection.Sat, 12 Sep 2026 09:30:00 GMTGaurav TiwariArchitecture & engineeringEvaluation & assuranceEvaluation evidenceRelease decisionsDecisions behind dependable agentshttps://softlander.co/insights/blog/evaluating-agent-readiness/https://softlander.co/insights/blog/evaluating-agent-readiness/A compelling demo leaves important questions open. Use a five-dimension rubric to examine the data, authority, evidence, ownership and people behind your next agent initiative.Sat, 12 Sep 2026 09:00:00 GMTGaurav TiwariGovernance & privacyStrategy & transformationAgent readinessOperating modelDecisions behind dependable agentshttps://softlander.co/insights/blog/permission-aware-retrieval-control-rag-diagrams-omit/https://softlander.co/insights/blog/permission-aware-retrieval-control-rag-diagrams-omit/A retrieval system should establish the active principal and filter authorised resources before similarity ranking or model context assembly.Thu, 10 Sep 2026 08:30:00 GMTGaurav TiwariArchitecture & engineeringGovernance & privacyIdentity and accessMemory and contextPrivacy engineeringGovernable agentshttps://softlander.co/insights/blog/computer-use-agents-interface-threat-model/https://softlander.co/insights/blog/computer-use-agents-interface-threat-model/A computer-use agent treats pixels, accessibility data and interface state as both evidence and an action surface. Security must bind intent to effects beyond the screen.Wed, 09 Sep 2026 08:30:00 GMTGaurav TiwariAgent securityArchitecture & engineeringPrompt injectionSandboxingTool securityFrontier agentshttps://softlander.co/insights/blog/stop-choosing-ai-use-cases-start-defining-workflows/https://softlander.co/insights/blog/stop-choosing-ai-use-cases-start-defining-workflows/A list of AI ideas is not an operating plan. Define the work, decisions, authority and evidence before deciding whether an agent belongs in the workflow.Tue, 08 Sep 2026 11:00:00 GMTGaurav TiwariStrategy & transformationAgent readinessOperating modelProduct managementAgentic transformationhttps://softlander.co/insights/blog/cto-checklist-agent-production/https://softlander.co/insights/blog/cto-checklist-agent-production/Production readiness for an AI agent spans workflow requirements, system identity, evaluation, recovery and operations—not just a working model endpoint.Mon, 07 Sep 2026 09:00:00 GMTGaurav TiwariArchitecture & engineeringEvaluation & assuranceAgent architectureExecutive leadershipRelease decisionsLeadership and researchhttps://softlander.co/insights/blog/ai-agents-in-legal-operations/https://softlander.co/insights/blog/ai-agents-in-legal-operations/Legal agents can organize evidence and prepare work, but professional judgement, client authority and verified legal positions cannot be delegated by interface.Fri, 04 Sep 2026 08:00:00 GMTGaurav TiwariAgent securityGovernance & privacyAgent authorityHuman oversightIndustry applicationsPrivacy engineeringAgents in practicehttps://softlander.co/insights/blog/human-handoff-preserves-context-authority/https://softlander.co/insights/blog/human-handoff-preserves-context-authority/A safe handoff transfers the decision, evidence, current state and bounded authority—not a transcript dump or a vague request for someone to take over.Thu, 03 Sep 2026 08:00:00 GMTGaurav TiwariArchitecture & engineeringGovernance & privacyAgent authorityHuman oversightOrchestrationAgent systemshttps://softlander.co/insights/blog/when-open-source-agent-infrastructure-is-responsible-choice/https://softlander.co/insights/blog/when-open-source-agent-infrastructure-is-responsible-choice/Openness can improve inspectability, control and portability, but only when an organisation can own the supply chain, operation and evidence it inherits.Wed, 02 Sep 2026 08:00:00 GMTGaurav TiwariArchitecture & engineeringGovernance & privacyAgent architectureGovernance controlsProcurementFrontier agentshttps://softlander.co/insights/blog/ai-feature-agent-and-agentic-system/https://softlander.co/insights/blog/ai-feature-agent-and-agentic-system/Classify AI by the action path it creates, not the label on the product. The distinction determines ownership, controls, evaluation and release evidence.Tue, 01 Sep 2026 11:00:00 GMTGaurav TiwariArchitecture & engineeringStrategy & transformationAgent architectureAgent authorityOperating modelFrontier agentshttps://softlander.co/insights/blog/failure-ladder-for-retrieval-augmented-agents/https://softlander.co/insights/blog/failure-ladder-for-retrieval-augmented-agents/A wrong answer is the final symptom of failures that may begin with source coverage, access control, parsing, retrieval, context assembly, generation or action.Mon, 31 Aug 2026 07:30:00 GMTGaurav TiwariArchitecture & engineeringEvaluation & assuranceEvaluation designFailure recoveryMemory and contextAgent systemshttps://softlander.co/insights/blog/agent-governance-review-triggers/https://softlander.co/insights/blog/agent-governance-review-triggers/Review should follow material changes in purpose, authority, data, users, dependencies, evidence or incidents—not an arbitrary calendar alone.Fri, 28 Aug 2026 13:00:00 GMTGaurav TiwariEvaluation & assuranceGovernance & privacyContinuous assuranceEvaluation evidenceGovernance controlsGovernable agentshttps://softlander.co/insights/blog/inventory-before-ai-governance/https://softlander.co/insights/blog/inventory-before-ai-governance/An AI inventory becomes useful when it identifies deployed systems, their context, owners, dependencies and review triggers—not merely a list of model vendors.Thu, 27 Aug 2026 09:00:00 GMTGaurav TiwariGovernance & privacyEvaluation evidenceGovernance controlsGovernable agentshttps://softlander.co/insights/blog/write-an-rfp-for-agentic-ai/https://softlander.co/insights/blog/write-an-rfp-for-agentic-ai/Specify the workflow, evidence, responsibility and exit conditions suppliers must meet, then evaluate them through representative scenarios.Wed, 26 Aug 2026 10:00:00 GMTGaurav TiwariStrategy & transformationEvaluation designProcurementRelease decisionsProduct and valuehttps://softlander.co/insights/blog/credentials-that-expire-before-trust/https://softlander.co/insights/blog/credentials-that-expire-before-trust/Agent credentials should be shorter-lived than the assumptions that justify them. Bind them to a workload, audience, task and action, then prove renewal and revocation work.Tue, 25 Aug 2026 09:00:00 GMTGaurav TiwariAgent securityArchitecture & engineeringIdentity and accessIncident readinessTool securitySecure agentshttps://softlander.co/insights/blog/do-you-need-agent-centre-of-excellence/https://softlander.co/insights/blog/do-you-need-agent-centre-of-excellence/A centre can concentrate scarce expertise and reusable controls, or become an approval queue detached from real work. Choose the mandate before the structure.Mon, 24 Aug 2026 08:30:00 GMTGaurav TiwariStrategy & transformationGovernance controlsOperating modelOrganisational changeAgentic transformationhttps://softlander.co/insights/blog/smaller-models-strengthen-agent-architecture/https://softlander.co/insights/blog/smaller-models-strengthen-agent-architecture/Smaller models are not merely budget substitutes. Used behind explicit routing and escalation rules, they can improve latency, locality and architectural control.Fri, 21 Aug 2026 09:30:00 GMTGaurav TiwariArchitecture & engineeringImpact measurementAgent architectureOrchestrationValue measurementFrontier agentshttps://softlander.co/insights/blog/golden-traces-are-not-gold-standards/https://softlander.co/insights/blog/golden-traces-are-not-gold-standards/A reference trace can explain one acceptable journey. Treating it as the only correct path can punish valid behaviour and conceal unsafe alternatives.Thu, 20 Aug 2026 13:00:00 GMTGaurav TiwariArchitecture & engineeringEvaluation & assuranceAgent reliabilityEvaluation designEvaluation evidenceRelease evidencehttps://softlander.co/insights/blog/quarterly-watchlist-agentic-ai-leaders/https://softlander.co/insights/blog/quarterly-watchlist-agentic-ai-leaders/A useful quarterly watchlist tracks changes that could alter real decisions: system authority, evidence, incidents, standards, providers and organisational readiness.Wed, 19 Aug 2026 10:00:00 GMTGaurav TiwariGovernance & privacyStrategy & transformationContinuous assuranceExecutive leadershipOperating modelLeadership and researchhttps://softlander.co/insights/blog/evidence-for-an-ai-security-incident/https://softlander.co/insights/blog/evidence-for-an-ai-security-incident/Investigators need more than chat transcripts. Preserve identity, instructions, retrieved context, model and tool versions, policy decisions, credentials, network events and actual state changes.Tue, 18 Aug 2026 09:00:00 GMTGaurav TiwariAgent securityGovernance & privacyIncident readinessObservabilityPrompt injectionSecure agentshttps://softlander.co/insights/blog/redesigning-roles-around-human-agent-collaboration/https://softlander.co/insights/blog/redesigning-roles-around-human-agent-collaboration/Do not divide work into “human” and “AI” columns. Redesign tasks, decisions, authority, handoffs and learning so each role remains operable and accountable.Mon, 17 Aug 2026 09:00:00 GMTGaurav TiwariImpact measurementStrategy & transformationHuman oversightOperating modelOrganisational changeAgentic transformationhttps://softlander.co/insights/blog/why-ai-pilots-become-permanent-experiments/https://softlander.co/insights/blog/why-ai-pilots-become-permanent-experiments/A successful demonstration does not create a live service. Pilots stall when they test capability without resolving ownership, integration, economics and operating evidence.Fri, 14 Aug 2026 13:00:00 GMTGaurav TiwariStrategy & transformationAgent readinessOperating modelValue measurementAgentic transformationhttps://softlander.co/insights/blog/certificate-expiry-outages-are-a-design-failure/https://softlander.co/insights/blog/certificate-expiry-outages-are-a-design-failure/A certificate that expires in production was not forgotten by an individual. It was issued by a process that did not include renewal, into a system that had no way to reload it.Thu, 13 Aug 2026 08:00:00 GMTGaurav TiwariCryptographic engineeringKey managementPublic-key infrastructureCryptography in productionhttps://softlander.co/insights/blog/ai-agent-incident-tabletop-exercise/https://softlander.co/insights/blog/ai-agent-incident-tabletop-exercise/A useful tabletop follows one plausible agent failure through detection, containment, evidence, business recovery and accountable return to service.Wed, 12 Aug 2026 13:00:00 GMTGaurav TiwariAgent securityGovernance & privacyExecutive leadershipFailure recoveryIncident readinessLeadership and researchhttps://softlander.co/insights/blog/estimate-the-cost-of-an-agent-failure/https://softlander.co/insights/blog/estimate-the-cost-of-an-agent-failure/Model concrete failure scenarios across detection, containment, correction, disruption and consequence, then use ranges to choose controls.Tue, 11 Aug 2026 11:00:00 GMTGaurav TiwariEvaluation & assuranceImpact measurementEvaluation evidenceRelease decisionsValue measurementProduct and valuehttps://softlander.co/insights/blog/testing-agent-network-egress/https://softlander.co/insights/blog/testing-agent-network-egress/“Internet access” is too broad a permission for an agent. Define destinations, protocols, data-release rules and failure behaviour, then test bypasses against the deployed network path.Mon, 10 Aug 2026 11:00:00 GMTGaurav TiwariAgent securityArchitecture & engineeringIncident readinessSandboxingTool securitySecure agentshttps://softlander.co/insights/blog/board-questions-ai-agents-authority/https://softlander.co/insights/blog/board-questions-ai-agents-authority/Boards do not need to review prompts or model scores. They need a clear account of what each agent may do, whose mandate it carries and how control is recovered.Thu, 06 Aug 2026 08:30:00 GMTGaurav TiwariGovernance & privacyStrategy & transformationAgent authorityExecutive leadershipGovernance controlsLeadership and researchhttps://softlander.co/insights/blog/agentic-ai-in-manufacturing-maintenance/https://softlander.co/insights/blog/agentic-ai-in-manufacturing-maintenance/Maintenance agents can turn condition signals into coordinated work, but plant safety and control demand bounded action and field verification.Wed, 05 Aug 2026 10:00:00 GMTGaurav TiwariAgent securityArchitecture & engineeringAgent authorityFailure recoveryIndustry applicationsTool securityAgents in practicehttps://softlander.co/insights/blog/design-joint-ai-evaluation-lab/https://softlander.co/insights/blog/design-joint-ai-evaluation-lab/A joint lab should be a governed way to create decision evidence across organisations, with shared methods but explicit ownership of systems, data and conclusions.Tue, 04 Aug 2026 08:30:00 GMTGaurav TiwariArchitecture & engineeringEvaluation & assuranceClient collaborationEvaluation evidencePartnershipsResearch methodsEcosystem and experiencehttps://softlander.co/insights/blog/public-sector-buyers-agent-supplier/https://softlander.co/insights/blog/public-sector-buyers-agent-supplier/Buy evidence, controllability and lifecycle obligations—not a confident demo or an unrestricted promise of autonomy.Mon, 03 Aug 2026 08:00:00 GMTGaurav TiwariGovernance & privacyStrategy & transformationEvaluation evidenceGovernance controlsIndustry applicationsProcurementAgents in practicehttps://softlander.co/insights/blog/durable-agent-workflows-need-explicit-state/https://softlander.co/insights/blog/durable-agent-workflows-need-explicit-state/Long-running agent work cannot rely on a transcript and hopeful retries. Explicit state makes progress, authority, recovery and completion inspectable across failures.Fri, 31 Jul 2026 08:30:00 GMTGaurav TiwariArchitecture & engineeringAgent architectureFailure recoveryOrchestrationAgent systemshttps://softlander.co/insights/blog/product-acceptance-criteria-for-non-deterministic-systems/https://softlander.co/insights/blog/product-acceptance-criteria-for-non-deterministic-systems/Accept an agent against outcome distributions, critical invariants and recovery evidence—not one scripted response or an average score.Thu, 30 Jul 2026 08:00:00 GMTGaurav TiwariEvaluation & assuranceEvaluation designProduct managementRelease decisionsProduct and valuehttps://softlander.co/insights/blog/hidden-total-cost-of-operating-an-ai-agent/https://softlander.co/insights/blog/hidden-total-cost-of-operating-an-ai-agent/Model usage is one line in a wider cost stack that includes integration, evaluation, control, human work, incidents, change and exit.Wed, 29 Jul 2026 13:00:00 GMTGaurav TiwariImpact measurementStrategy & transformationProcurementProduct managementValue measurementProduct and valuehttps://softlander.co/insights/blog/cryptographic-agility-is-an-interface-problem/https://softlander.co/insights/blog/cryptographic-agility-is-an-interface-problem/Agility is not a property of the algorithms you chose. It is a property of how your code, your data formats and your protocols name them—and whether anything can change without a coordinated rewrite.Tue, 28 Jul 2026 10:00:00 GMTGaurav TiwariCryptographic engineeringApplied cryptographyCryptographic agilityCryptography in productionhttps://softlander.co/insights/blog/ai-agent-decision-reviewable-artifacts/https://softlander.co/insights/blog/ai-agent-decision-reviewable-artifacts/Reviewable agent decisions depend on a small chain of artefacts connecting workflow purpose, authority, configuration, evidence, findings, controls and ownership.Mon, 27 Jul 2026 13:00:00 GMTGaurav TiwariEvaluation & assuranceGovernance & privacyEvaluation evidenceExecutive leadershipGovernance controlsLeadership and researchhttps://softlander.co/insights/blog/no-agent-test-ordinary-automation/https://softlander.co/insights/blog/no-agent-test-ordinary-automation/Agents earn their complexity only when a workflow needs bounded adaptation. A disciplined no-agent test can reveal when rules, search or ordinary software is safer and cheaper.Fri, 24 Jul 2026 13:00:00 GMTGaurav TiwariArchitecture & engineeringStrategy & transformationAgent readinessProduct managementValue measurementAgentic transformationhttps://softlander.co/insights/blog/least-privilege-for-ai-agents/https://softlander.co/insights/blog/least-privilege-for-ai-agents/Least privilege for agents is not a smaller role assigned once. It is authority narrowed by identity, task, resource, action, time and consequence at every tool call.Thu, 23 Jul 2026 10:00:00 GMTGaurav TiwariAgent securityArchitecture & engineeringIdentity and accessTool securitySecure agentshttps://softlander.co/insights/blog/can-your-agent-truly-forget/https://softlander.co/insights/blog/can-your-agent-truly-forget/Deleting a chat is not the same as removing personal data from memory, retrieval, logs, caches, evaluations, backups and learned models.Wed, 22 Jul 2026 11:00:00 GMTGaurav TiwariArchitecture & engineeringGovernance & privacyEvaluation evidenceMemory and contextPrivacy engineeringGovernable agentshttps://softlander.co/insights/blog/authorization-is-where-api-breaches-begin/https://softlander.co/insights/blog/authorization-is-where-api-breaches-begin/Most APIs authenticate well and authorise badly. The defects that matter are object-level and function-level access rules scattered across handlers, where no single reviewer can see the whole rule.Tue, 21 Jul 2026 08:00:00 GMTGaurav TiwariSecurity engineeringAPI securityIdentity and accessSecure by designhttps://softlander.co/insights/blog/you-cannot-migrate-cryptography-you-cannot-find/https://softlander.co/insights/blog/you-cannot-migrate-cryptography-you-cannot-find/Every post-quantum migration plan begins with an inventory, and most inventories stall because the team looks for algorithms instead of looking for the places algorithms are decided.Mon, 20 Jul 2026 10:00:00 GMTGaurav TiwariCryptographic engineeringCryptographic agilityPost-quantum cryptographyCryptography in productionhttps://softlander.co/insights/blog/evaluate-software-engineering-agent-real-repository-work/https://softlander.co/insights/blog/evaluate-software-engineering-agent-real-repository-work/A passing patch is only the start. Repository evaluation should test understanding, change quality, security, reviewability and recovery.Fri, 17 Jul 2026 09:00:00 GMTGaurav TiwariAgent securityArchitecture & engineeringEvaluation & assuranceAgent reliabilityEvaluation designIndustry applicationsTool securityAgents in practicehttps://softlander.co/insights/blog/transformation-roadmap-built-around-decisions/https://softlander.co/insights/blog/transformation-roadmap-built-around-decisions/Model launches are not transformation milestones. Organize the roadmap around workflow, evidence, authority, operating and expansion decisions.Thu, 16 Jul 2026 08:00:00 GMTGaurav TiwariStrategy & transformationOperating modelOrganisational changeProduct managementAgentic transformationhttps://softlander.co/insights/blog/agent-protocols-new-control-surface/https://softlander.co/insights/blog/agent-protocols-new-control-surface/Protocols make tools and agents interoperable, but they also standardise discovery, delegation and context exchange. Those interfaces need explicit trust and policy controls.Wed, 15 Jul 2026 08:00:00 GMTGaurav TiwariAgent securityArchitecture & engineeringAgent protocolsIdentity and accessTool securityFrontier agentshttps://softlander.co/insights/blog/measure-agent-action-not-answer/https://softlander.co/insights/blog/measure-agent-action-not-answer/A convincing message can conceal a failed or unauthorised operation. Agent evaluation must inspect tool calls, external state and control decisions.Tue, 14 Jul 2026 08:00:00 GMTGaurav TiwariArchitecture & engineeringEvaluation & assuranceAgent authorityEvaluation designEvaluation evidenceRelease evidencehttps://softlander.co/insights/blog/write-ai-success-story-that-survives-scrutiny/https://softlander.co/insights/blog/write-ai-success-story-that-survives-scrutiny/A credible AI success story separates what was delivered from what changed, defines the baseline, shows the contribution chain and preserves uncertainty and permission.Mon, 13 Jul 2026 09:30:00 GMTGaurav TiwariEvaluation & assuranceImpact measurementClient collaborationEvaluation evidenceResearch methodsValue measurementEcosystem and experiencehttps://softlander.co/insights/blog/collaborate-when-client-data-cannot-leave-environment/https://softlander.co/insights/blog/collaborate-when-client-data-cannot-leave-environment/Keep sensitive data under client control by moving approved methods and people to the environment, exporting only the minimum reviewed evidence needed for collaboration.Fri, 10 Jul 2026 13:00:00 GMTGaurav TiwariArchitecture & engineeringGovernance & privacyClient collaborationIdentity and accessPartnershipsPrivacy engineeringEcosystem and experiencehttps://softlander.co/insights/blog/post-quantum-readiness-starts-with-data-lifetime/https://softlander.co/insights/blog/post-quantum-readiness-starts-with-data-lifetime/The question that sequences a post-quantum programme is not which algorithm to adopt. It is which of your data, recorded by an adversary today, would still matter when it can be decrypted.Thu, 09 Jul 2026 08:00:00 GMTGaurav TiwariCryptographic engineeringApplied cryptographyPost-quantum cryptographyCryptography in productionhttps://softlander.co/insights/blog/evaluating-an-agent-orchestrator/https://softlander.co/insights/blog/evaluating-an-agent-orchestrator/An orchestrator should be judged by the workflow it controls: decomposition, routing, state, authority, recovery, verification and stopping—not by a polished final response.Wed, 08 Jul 2026 08:30:00 GMTGaurav TiwariArchitecture & engineeringEvaluation & assuranceAgent reliabilityEvaluation designOrchestrationAgent systemshttps://softlander.co/insights/blog/human-oversight-is-a-designed-control/https://softlander.co/insights/blog/human-oversight-is-a-designed-control/Effective oversight gives a prepared person the signal, authority, time and interface needed to prevent or limit a defined AI failure.Tue, 07 Jul 2026 11:00:00 GMTGaurav TiwariEvaluation & assuranceGovernance & privacyEvaluation evidenceGovernance controlsHuman oversightGovernable agentshttps://softlander.co/insights/blog/continuous-evaluation-without-alert-theatre/https://softlander.co/insights/blog/continuous-evaluation-without-alert-theatre/Run evaluations often, but interrupt people only when a signal maps to user consequence, a decision owner and a tested response.Mon, 06 Jul 2026 08:00:00 GMTGaurav TiwariArchitecture & engineeringEvaluation & assuranceAgent reliabilityContinuous assuranceObservabilityRelease evidencehttps://softlander.co/insights/blog/security-review-for-an-mcp-server/https://softlander.co/insights/blog/security-review-for-an-mcp-server/Review an MCP server as a privileged protocol adapter: verify tool semantics, OAuth boundaries, token audience, downstream delegation, input handling, isolation, logging and consent.Thu, 02 Jul 2026 13:00:00 GMTGaurav TiwariAgent securityArchitecture & engineeringAgent protocolsIdentity and accessTool securitySecure agentshttps://softlander.co/insights/blog/threat-model-that-survives-architecture-review/https://softlander.co/insights/blog/threat-model-that-survives-architecture-review/Most threat models are produced once, filed, and never consulted again. A model earns its place when every ranked attack path is attached to an architectural decision someone has to make.Wed, 01 Jul 2026 09:30:00 GMTGaurav TiwariSecurity engineeringSecure developmentThreat modellingSecure by designhttps://softlander.co/insights/blog/which-security-gates-belong-in-the-pipeline/https://softlander.co/insights/blog/which-security-gates-belong-in-the-pipeline/Adding every available scanner to continuous integration produces a slow pipeline and a culture of overrides. Choose gates by what they can decide reliably, and move the rest out of the blocking path.Tue, 30 Jun 2026 09:00:00 GMTGaurav TiwariSecurity engineeringSecure developmentVulnerability remediationSecure by designhttps://softlander.co/insights/blog/your-dependency-graph-is-attack-surface/https://softlander.co/insights/blog/your-dependency-graph-is-attack-surface/A software bill of materials tells you what you ship. Controlling supply-chain risk needs three further things: provenance for the build, a policy for adoption, and a rehearsed route to replace a component quickly.Mon, 29 Jun 2026 09:00:00 GMTGaurav TiwariSecurity engineeringSupply-chain securityVulnerability remediationSecure by designhttps://softlander.co/insights/blog/turning-ai-workshop-into-decision-artifact/https://softlander.co/insights/blog/turning-ai-workshop-into-decision-artifact/Design an AI workshop backwards from a named decision, then capture evidence, assumptions, disagreements, owners and triggers in an artefact that survives the meeting.Fri, 26 Jun 2026 09:30:00 GMTGaurav TiwariGovernance & privacyStrategy & transformationClient collaborationExecutive leadershipGovernance controlsOperating modelEcosystem and experiencehttps://softlander.co/insights/blog/product-leaders-demand-agent-evaluation/https://softlander.co/insights/blog/product-leaders-demand-agent-evaluation/Product leaders need evaluation evidence that connects workflow outcomes, user experience and policy behaviour to a release choice—not a leaderboard score.Thu, 25 Jun 2026 09:30:00 GMTGaurav TiwariEvaluation & assuranceImpact measurementEvaluation designProduct managementRelease decisionsLeadership and researchhttps://softlander.co/insights/blog/future-ai-assurance-continuous-not-fully-automatic/https://softlander.co/insights/blog/future-ai-assurance-continuous-not-fully-automatic/Agent systems change after release, so evidence must keep moving. Automation can collect and test signals; people remain responsible for judgement and action.Wed, 24 Jun 2026 11:00:00 GMTGaurav TiwariEvaluation & assuranceGovernance & privacyContinuous assuranceEvaluation evidenceHuman oversightFrontier agentshttps://softlander.co/insights/blog/mapping-an-ai-agent-attack-surface/https://softlander.co/insights/blog/mapping-an-ai-agent-attack-surface/A useful agent threat model follows data, instructions, identity, tools and state across trust boundaries. This method turns the map into test cases and control decisions.Tue, 23 Jun 2026 09:30:00 GMTGaurav TiwariAgent securityArchitecture & engineeringIdentity and accessPrompt injectionTool securitySecure agentshttps://softlander.co/insights/blog/what-belongs-on-an-ai-impact-dashboard/https://softlander.co/insights/blog/what-belongs-on-an-ai-impact-dashboard/Give decision owners a balanced view of outcomes, quality, human work, cost, risk and change—with every measure tied to an action.Mon, 22 Jun 2026 11:00:00 GMTGaurav TiwariImpact measurementEvaluation evidenceProduct managementValue measurementProduct and valuehttps://softlander.co/insights/blog/cryptography-is-an-architecture-decision/https://softlander.co/insights/blog/cryptography-is-an-architecture-decision/Choosing a well-regarded library settles very little. The decisions that determine whether cryptography protects anything are about where keys live, what a compromise reaches, and which threat you are actually addressing.Fri, 19 Jun 2026 09:00:00 GMTGaurav TiwariCryptographic engineeringApplied cryptographyKey managementCryptography in productionhttps://softlander.co/insights/blog/private-data-untrusted-content-external-actions/https://softlander.co/insights/blog/private-data-untrusted-content-external-actions/An agent becomes especially risky when it can read secrets, ingest attacker-controlled material and communicate or act outside the boundary. Design the workflow to break that combination.Thu, 18 Jun 2026 08:30:00 GMTGaurav TiwariAgent securityArchitecture & engineeringPrompt injectionSandboxingTool securitySecure agentshttps://softlander.co/insights/blog/what-belongs-in-a-post-quantum-migration-plan/https://softlander.co/insights/blog/what-belongs-in-a-post-quantum-migration-plan/A plan that lists algorithms and a target year is a statement of intent. A plan that survives review names sequenced workstreams, dependency owners, interim states and the decisions that are already overdue.Wed, 17 Jun 2026 13:00:00 GMTGaurav TiwariCryptographic engineeringCryptographic agilityPost-quantum cryptographyCryptography in productionhttps://softlander.co/insights/blog/where-ai-agents-fit-finance-operations/https://softlander.co/insights/blog/where-ai-agents-fit-finance-operations/Finance agents can prepare, reconcile and investigate, but posting, paying and attesting require deliberately separated authority.Tue, 16 Jun 2026 11:00:00 GMTGaurav TiwariGovernance & privacyStrategy & transformationAgent authorityGovernance controlsIndustry applicationsOperating modelAgents in practicehttps://softlander.co/insights/blog/conference-participation-evidence/https://softlander.co/insights/blog/conference-participation-evidence/Attendance, speaking and sponsorship are evidence of activity; useful conference claims require separate measures of learning, relationships, follow-through and resulting decisions.Mon, 15 Jun 2026 07:30:00 GMTGaurav TiwariImpact measurementStrategy & transformationPartnershipsResearch methodsValue measurementEcosystem and experiencehttps://softlander.co/insights/blog/ai-readiness-is-local-to-workflow/https://softlander.co/insights/blog/ai-readiness-is-local-to-workflow/Enterprise readiness scores hide the conditions that make one workflow viable and another reckless. Assess readiness where data, authority, people and operations meet.Fri, 12 Jun 2026 07:30:00 GMTGaurav TiwariGovernance & privacyStrategy & transformationAgent readinessGovernance controlsOperating modelAgentic transformationhttps://softlander.co/insights/blog/context-is-not-memory/https://softlander.co/insights/blog/context-is-not-memory/Context is the information available for one inference. Memory is a governed process for retaining, changing and recalling information across time. Confusing them creates silent risk.Thu, 11 Jun 2026 09:00:00 GMTGaurav TiwariArchitecture & engineeringGovernance & privacyAgent architectureMemory and contextPrivacy engineeringAgent systemshttps://softlander.co/insights/blog/cost-per-quality-adjusted-completion/https://softlander.co/insights/blog/cost-per-quality-adjusted-completion/Connect spending to accepted workflow outcomes by counting retries, review, correction and failures alongside model and platform charges.Wed, 10 Jun 2026 11:00:00 GMTGaurav TiwariEvaluation & assuranceImpact measurementEvaluation evidenceProduct managementValue measurementProduct and valuehttps://softlander.co/insights/blog/technical-ai-privacy-review-boundaries/https://softlander.co/insights/blog/technical-ai-privacy-review-boundaries/A technical review can trace data and test controls, but it cannot alone determine lawful basis, resolve policy choices or certify compliance.Tue, 09 Jun 2026 08:30:00 GMTGaurav TiwariEvaluation & assuranceGovernance & privacyEvaluation evidenceGovernance controlsPrivacy engineeringGovernable agentshttps://softlander.co/insights/blog/turn-red-team-findings-into-release-criteria/https://softlander.co/insights/blog/turn-red-team-findings-into-release-criteria/A dramatic exploit is a discovery, not a decision rule. Translate it into a threat condition, required control, repeatable regression and release consequence.Mon, 08 Jun 2026 07:30:00 GMTGaurav TiwariAgent securityEvaluation & assuranceEvaluation evidencePrompt injectionRelease decisionsRelease evidencehttps://softlander.co/insights/blog/responsible-autonomy-workshop/https://softlander.co/insights/blog/responsible-autonomy-workshop/Responsible autonomy becomes concrete when a cross-functional team maps one workflow’s actions, consequences, evidence and intervention choices.Fri, 05 Jun 2026 10:00:00 GMTGaurav TiwariGovernance & privacyStrategy & transformationAgent authorityExecutive leadershipHuman oversightLeadership and researchhttps://softlander.co/insights/blog/cisos-tool-using-ai-agents/https://softlander.co/insights/blog/cisos-tool-using-ai-agents/Tool-using agents join untrusted input, probabilistic decisions and privileged action. CISOs should govern that path as a security system, not a chatbot feature.Thu, 04 Jun 2026 08:00:00 GMTGaurav TiwariAgent securityArchitecture & engineeringExecutive leadershipPrompt injectionTool securityLeadership and researchhttps://softlander.co/insights/blog/qualify-ai-technology-partner-before-joint-offer/https://softlander.co/insights/blog/qualify-ai-technology-partner-before-joint-offer/Qualify a technology partner through evidence of fit, controls, delivery behaviour and commercial alignment before combining brands or promising a joint service.Wed, 03 Jun 2026 10:00:00 GMTGaurav TiwariGovernance & privacyStrategy & transformationGovernance controlsPartnershipsProcurementEcosystem and experiencehttps://softlander.co/insights/blog/build-buy-or-partner-for-an-ai-agent/https://softlander.co/insights/blog/build-buy-or-partner-for-an-ai-agent/Choose a sourcing path by separating the workflow advantage you must own from the capabilities you can responsibly obtain elsewhere.Tue, 02 Jun 2026 09:00:00 GMTGaurav TiwariImpact measurementStrategy & transformationProcurementProduct managementRelease decisionsProduct and valuehttps://softlander.co/insights/blog/evaluating-customer-support-agent-answer-to-action/https://softlander.co/insights/blog/evaluating-customer-support-agent-answer-to-action/A support agent should be judged on whether it resolves the right problem safely, not on whether its first answer sounds convincing.Mon, 01 Jun 2026 09:00:00 GMTGaurav TiwariArchitecture & engineeringEvaluation & assuranceAgent authorityEvaluation designHuman oversightIndustry applicationsAgents in practicehttps://softlander.co/insights/blog/encryption-at-rest-answers-fewer-questions/https://softlander.co/insights/blog/encryption-at-rest-answers-fewer-questions/Platform-managed encryption satisfies a procurement question and defends against one threat. Knowing which threat it covers—and which of your actual risks it leaves untouched—changes what you build next.Thu, 28 May 2026 07:30:00 GMTGaurav TiwariCryptographic engineeringApplied cryptographyPrivacy engineeringCryptography in productionhttps://softlander.co/insights/blog/agent-observability-answers-questions/https://softlander.co/insights/blog/agent-observability-answers-questions/Useful observability reconstructs a decision and its effects. Begin with incident, release and governance questions; then collect the smallest safe evidence that answers them.Wed, 27 May 2026 09:00:00 GMTGaurav TiwariArchitecture & engineeringEvaluation & assuranceContinuous assuranceIncident readinessObservabilityAgent systemshttps://softlander.co/insights/blog/confidence-for-agent-release-decision/https://softlander.co/insights/blog/confidence-for-agent-release-decision/Confidence is not a universal percentage. It is a judgement about evidence quality, uncertainty, consequence and the reversibility of the proposed release.Tue, 26 May 2026 10:00:00 GMTGaurav TiwariEvaluation & assuranceEvaluation designEvaluation evidenceRelease decisionsRelease evidencehttps://softlander.co/insights/blog/autonomy-classification-starts-with-consequences/https://softlander.co/insights/blog/autonomy-classification-starts-with-consequences/Classify an AI agent by what it can affect, the authority it can exercise and how quickly people can recover—not by how autonomous a demo appears.Mon, 25 May 2026 08:30:00 GMTGaurav TiwariArchitecture & engineeringGovernance & privacyAgent authorityGovernance controlsHuman oversightGovernable agentshttps://softlander.co/insights/blog/what-makes-ai-assurance-partnership-useful/https://softlander.co/insights/blog/what-makes-ai-assurance-partnership-useful/A useful assurance partner improves the quality of a specific AI decision while leaving ownership, evidence and operational accountability with the organisation.Fri, 22 May 2026 11:00:00 GMTGaurav TiwariEvaluation & assuranceStrategy & transformationClient collaborationEvaluation evidencePartnershipsEcosystem and experiencehttps://softlander.co/insights/blog/when-not-to-design-a-security-protocol/https://softlander.co/insights/blog/when-not-to-design-a-security-protocol/Almost every custom security protocol should have been a profile of an existing one. The few that are justified share a specific set of constraints - and a review process that assumes the first design is wrong.Thu, 21 May 2026 08:30:00 GMTGaurav TiwariSecurity engineeringApplied cryptographyProtocol designSecure by designhttps://softlander.co/insights/blog/good-client-controlled-ai-evaluation/https://softlander.co/insights/blog/good-client-controlled-ai-evaluation/Client control means owning the decision standard, representative cases, evidence access and release authority—not forcing every test to be built or run internally.Wed, 20 May 2026 07:30:00 GMTGaurav TiwariEvaluation & assuranceGovernance & privacyClient collaborationEvaluation evidenceGovernance controlsEcosystem and experiencehttps://softlander.co/insights/blog/ai-productivity-claims-need-a-baseline/https://softlander.co/insights/blog/ai-productivity-claims-need-a-baseline/Time saved is not an outcome until the original work, quality, case mix and destination of released capacity are measured on comparable terms.Tue, 19 May 2026 11:00:00 GMTGaurav TiwariImpact measurementEvaluation designProduct managementValue measurementProduct and valuehttps://softlander.co/insights/blog/authority-boundaries-for-procurement-agents/https://softlander.co/insights/blog/authority-boundaries-for-procurement-agents/A procurement agent may discover, compare and prepare, but supplier commitments require explicit mandates, separation and evidence.Mon, 18 May 2026 11:00:00 GMTGaurav TiwariAgent securityGovernance & privacyAgent authorityIdentity and accessIndustry applicationsProcurementAgents in practicehttps://softlander.co/insights/blog/portable-evaluations-protect-against-platform-lock-in/https://softlander.co/insights/blog/portable-evaluations-protect-against-platform-lock-in/Keep workflow cases, graders and decision rules separate from the platform so alternatives can be compared without rebuilding the standard of proof.Fri, 15 May 2026 08:00:00 GMTGaurav TiwariEvaluation & assuranceStrategy & transformationEvaluation designEvaluation evidenceProcurementProduct and valuehttps://softlander.co/insights/blog/where-do-your-keys-actually-live/https://softlander.co/insights/blog/where-do-your-keys-actually-live/Most organisations can describe their intended key management and cannot answer the question for a specific key. The gap between the two is where incidents and failed audits come from.Thu, 14 May 2026 11:00:00 GMTGaurav TiwariCryptographic engineeringApplied cryptographyKey managementCryptography in productionhttps://softlander.co/insights/blog/grade-agent-evaluation-findings/https://softlander.co/insights/blog/grade-agent-evaluation-findings/Grade findings by consequence, reach, reproducibility and control strength, then keep that technical judgement separate from the release decision.Wed, 13 May 2026 07:30:00 GMTGaurav TiwariEvaluation & assuranceEvaluation designEvaluation evidenceRelease decisionsRelease evidencehttps://softlander.co/insights/blog/when-agent-evaluation-dataset-becomes-stale/https://softlander.co/insights/blog/when-agent-evaluation-dataset-becomes-stale/Age is only one signal. A dataset becomes stale when its cases, labels or conditions no longer support the decision the evaluation is meant to inform.Tue, 12 May 2026 07:30:00 GMTGaurav TiwariEvaluation & assuranceContinuous assuranceEvaluation designEvaluation evidenceRelease evidencehttps://softlander.co/insights/blog/anatomy-of-a-tool-using-ai-agent/https://softlander.co/insights/blog/anatomy-of-a-tool-using-ai-agent/A tool-using agent is a controlled system around a model, not a model with an API key. Map its runtime layers before deciding what to test, trust or release.Mon, 11 May 2026 08:00:00 GMTGaurav TiwariArchitecture & engineeringGovernance & privacyAgent architectureAgent authorityOrchestrationAgent systemshttps://softlander.co/insights/blog/evaluating-model-routing-decisions/https://softlander.co/insights/blog/evaluating-model-routing-decisions/A model router is a policy for allocating quality, cost, latency, privacy and failure risk. Evaluate whether each route serves the workflow—not whether routing is fashionable.Fri, 08 May 2026 08:30:00 GMTGaurav TiwariArchitecture & engineeringEvaluation & assuranceAgent reliabilityEvaluation designOrchestrationAgent systemshttps://softlander.co/insights/blog/multi-agent-handoff-failures/https://softlander.co/insights/blog/multi-agent-handoff-failures/Multi-agent systems fail between workers when goals, evidence, state, authority and completion semantics do not survive the transfer. Test the protocol, not just each agent.Thu, 07 May 2026 09:00:00 GMTGaurav TiwariArchitecture & engineeringEvaluation & assuranceAgent protocolsFailure recoveryOrchestrationAgent systemshttps://softlander.co/insights/blog/vendor-neutral-does-not-mean-anti-vendor/https://softlander.co/insights/blog/vendor-neutral-does-not-mean-anti-vendor/Neutral advice compares dependencies and evidence against the client's decision; it does not pretend every platform is interchangeable or that vendors add no value.Wed, 06 May 2026 11:00:00 GMTGaurav TiwariArchitecture & engineeringStrategy & transformationClient collaborationPartnershipsProcurementEcosystem and experiencehttps://softlander.co/insights/blog/what-a-security-release-decision-rests-on/https://softlander.co/insights/blog/what-a-security-release-decision-rests-on/"The scans are green" is not a security release decision. A defensible one states what was tested, what is known to be wrong, who accepted it, and what would change the answer.Tue, 05 May 2026 08:00:00 GMTGaurav TiwariSecurity engineeringRelease assuranceVulnerability remediationSecure by designhttps://softlander.co/insights/blog/ai-agent-release-evidence-pack/https://softlander.co/insights/blog/ai-agent-release-evidence-pack/A useful release pack connects the intended workflow, evaluated behaviour, unresolved risk and operating controls to one accountable decision.Mon, 04 May 2026 10:00:00 GMTGaurav TiwariEvaluation & assuranceAgent reliabilityEvaluation evidenceRelease decisionsRelease evidencehttps://softlander.co/insights/blog/sandboxing-and-authorization/https://softlander.co/insights/blog/sandboxing-and-authorization/A sandbox limits what compromised execution can reach. Authorisation decides what an authenticated actor may do. Secure agents need both, tested independently and together.Fri, 01 May 2026 09:30:00 GMTGaurav TiwariAgent securityArchitecture & engineeringIdentity and accessSandboxingTool securitySecure agentshttps://softlander.co/insights/blog/what-a-polished-agent-demo-does-not-tell-you/https://softlander.co/insights/blog/what-a-polished-agent-demo-does-not-tell-you/A demonstration shows one possible path. A buying decision needs evidence about representative work, failure, authority, recovery and operation.Thu, 30 Apr 2026 09:30:00 GMTGaurav TiwariEvaluation & assuranceStrategy & transformationEvaluation evidenceProcurementProduct managementProduct and valuehttps://softlander.co/insights/blog/idempotency-for-ai-agents/https://softlander.co/insights/blog/idempotency-for-ai-agents/Retries, duplicated messages and uncertain tool outcomes can multiply an agent’s effects. Idempotency turns a logical action into one controlled business operation.Wed, 29 Apr 2026 10:00:00 GMTGaurav TiwariArchitecture & engineeringEvaluation & assuranceAgent reliabilityFailure recoveryOrchestrationAgent systemshttps://softlander.co/insights/blog/hybrid-key-exchange-what-it-protects-and-costs/https://softlander.co/insights/blog/hybrid-key-exchange-what-it-protects-and-costs/Combining a classical and a post-quantum key exchange is now the default for TLS. Understanding why the combination is used—and what it adds to every handshake—makes the deployment decision straightforward.Tue, 28 Apr 2026 07:30:00 GMTGaurav TiwariCryptographic engineeringPost-quantum cryptographyProtocol designCryptography in productionhttps://softlander.co/insights/blog/independent-agent-benchmark/https://softlander.co/insights/blog/independent-agent-benchmark/A durable agent benchmark needs a declared construct, protected cases, reproducible execution, uncertainty, governance and limits on the claims its scores can support.Mon, 27 Apr 2026 10:00:00 GMTGaurav TiwariEvaluation & assuranceEvaluation designEvaluation evidenceResearch methodsLeadership and researchhttps://softlander.co/insights/blog/responsibility-map-every-ai-partnership-needs/https://softlander.co/insights/blog/responsibility-map-every-ai-partnership-needs/A useful responsibility map connects each AI decision and artefact to a named accountable owner, a performing role, required contributors and evidence of completion.Thu, 23 Apr 2026 08:30:00 GMTGaurav TiwariGovernance & privacyStrategy & transformationClient collaborationGovernance controlsOperating modelPartnershipsEcosystem and experiencehttps://softlander.co/insights/blog/secure-boot-is-a-key-management-decision/https://softlander.co/insights/blog/secure-boot-is-a-key-management-decision/Teams treat secure boot as a firmware feature and discover years later that the hard part was never verification. It was deciding who holds the signing key, how it rotates, and what happens when it is lost.Wed, 22 Apr 2026 10:00:00 GMTGaurav TiwariSecurity engineeringKey managementSecure boot and updatesSecure by designhttps://softlander.co/insights/blog/prioritize-enterprise-agent-opportunities/https://softlander.co/insights/blog/prioritize-enterprise-agent-opportunities/Rank agent opportunities by decision value, workflow evidence, feasibility, consequence and reuse—not by enthusiasm, executive volume or demo polish.Tue, 21 Apr 2026 07:30:00 GMTGaurav TiwariImpact measurementStrategy & transformationAgent readinessProduct managementValue measurementAgentic transformationhttps://softlander.co/insights/blog/evaluate-agent-works-for-days/https://softlander.co/insights/blog/evaluate-agent-works-for-days/Long-running agents accumulate state, dependencies and chances to drift. Evaluation must test checkpoints, recovery, changing conditions and complete outcomes across time.Mon, 20 Apr 2026 08:00:00 GMTGaurav TiwariArchitecture & engineeringEvaluation & assuranceEvaluation designFailure recoveryLong-lived agentsFrontier agentshttps://softlander.co/insights/blog/secure-code-review-that-changes-the-code/https://softlander.co/insights/blog/secure-code-review-that-changes-the-code/A review that delivers a ranked findings report usually changes nothing. A review that delivers a patch, a test that fails before it and a pattern the team can apply changes the codebase.Fri, 17 Apr 2026 08:30:00 GMTGaurav TiwariSecurity engineeringSecure developmentVulnerability remediationSecure by designhttps://softlander.co/insights/blog/design-agent-evaluations-backwards-from-decision/https://softlander.co/insights/blog/design-agent-evaluations-backwards-from-decision/Start with the choice an owner must make, then derive claims, evidence, cases and measures. The result is smaller, sharper and easier to defend.Thu, 16 Apr 2026 10:00:00 GMTGaurav TiwariEvaluation & assuranceEvaluation designEvaluation evidenceRelease decisionsRelease evidencehttps://softlander.co/insights/blog/build-ai-evidence-register/https://softlander.co/insights/blog/build-ai-evidence-register/An evidence register should connect each claim and requirement to versioned artefacts, review decisions, known gaps and the conditions under which evidence expires.Wed, 15 Apr 2026 08:30:00 GMTGaurav TiwariEvaluation & assuranceGovernance & privacyEvaluation evidenceGovernance controlsGovernable agentshttps://softlander.co/insights/blog/autonomy-spectrum-actions-not-product-label/https://softlander.co/insights/blog/autonomy-spectrum-actions-not-product-label/Describe autonomy operation by operation. A multidimensional action map gives leaders a better basis for limits, oversight and release evidence than one marketing label.Tue, 14 Apr 2026 09:30:00 GMTGaurav TiwariArchitecture & engineeringGovernance & privacyAgent authorityGovernance controlsHuman oversightFrontier agentshttps://softlander.co/insights/blog/data-minimization-for-context-hungry-agents/https://softlander.co/insights/blog/data-minimization-for-context-hungry-agents/Better agent context is not the largest possible context. It is the smallest authorised, relevant and timely data set that supports the task.Mon, 13 Apr 2026 09:30:00 GMTGaurav TiwariArchitecture & engineeringGovernance & privacyIdentity and accessMemory and contextPrivacy engineeringGovernable agentshttps://softlander.co/insights/blog/when-ai-agent-can-modify-itself/https://softlander.co/insights/blog/when-ai-agent-can-modify-itself/Self-modification turns runtime learning into controlled software change. Separate what may adapt, require evidence for promotion and preserve a recoverable baseline.Fri, 10 Apr 2026 11:00:00 GMTGaurav TiwariArchitecture & engineeringEvaluation & assuranceGovernance & privacyContinuous assuranceGovernance controlsLong-lived agentsFrontier agentshttps://softlander.co/insights/blog/who-owns-ai-agent-after-launch/https://softlander.co/insights/blog/who-owns-ai-agent-after-launch/Live agents cut across product, operations, risk and technology. One accountable service owner needs explicit decision rights supported by named control owners.Thu, 09 Apr 2026 13:00:00 GMTGaurav TiwariGovernance & privacyStrategy & transformationGovernance controlsOperating modelProduct managementAgentic transformationhttps://softlander.co/insights/blog/testing-logistics-agents-under-disruption/https://softlander.co/insights/blog/testing-logistics-agents-under-disruption/A logistics agent earns trust when data becomes late, routes close and plans conflict—not when everything follows the timetable.Wed, 08 Apr 2026 08:00:00 GMTGaurav TiwariArchitecture & engineeringEvaluation & assuranceEvaluation designFailure recoveryIndustry applicationsOrchestrationAgents in practicehttps://softlander.co/insights/blog/testing-an-agent-kill-switch/https://softlander.co/insights/blog/testing-an-agent-kill-switch/Stopping one process may leave tools, queued actions, delegated agents and credentials active. Define the safe state, control every action path and rehearse the stop under load.Tue, 07 Apr 2026 11:00:00 GMTGaurav TiwariAgent securityGovernance & privacyAgent reliabilityIdentity and accessIncident readinessSecure agentshttps://softlander.co/insights/blog/signatures-are-the-harder-half-of-post-quantum/https://softlander.co/insights/blog/signatures-are-the-harder-half-of-post-quantum/Key exchange migration is largely a configuration change. Signature migration runs into certificate sizes, fixed-width image headers, verifiers you cannot update and archives that must stay verifiable for decades.Mon, 06 Apr 2026 10:00:00 GMTGaurav TiwariCryptographic engineeringPost-quantum cryptographyPublic-key infrastructureCryptography in productionhttps://softlander.co/insights/blog/audit-committees-discuss-ai/https://softlander.co/insights/blog/audit-committees-discuss-ai/Audit committees can oversee AI through ownership, control evidence, reporting integrity and assurance scope without turning meetings into model-architecture tutorials.Fri, 03 Apr 2026 08:00:00 GMTGaurav TiwariGovernance & privacyEvaluation evidenceExecutive leadershipGovernance controlsLeadership and researchhttps://softlander.co/insights/blog/test-agent-recovery/https://softlander.co/insights/blog/test-agent-recovery/Inject realistic faults, verify state and judge the recovery path. A retry that finishes the task can still duplicate work or hide uncertainty.Thu, 02 Apr 2026 08:30:00 GMTGaurav TiwariArchitecture & engineeringEvaluation & assuranceAgent reliabilityEvaluation designFailure recoveryRelease evidencehttps://softlander.co/insights/blog/healthcare-administrative-agents-strict-boundaries/https://softlander.co/insights/blog/healthcare-administrative-agents-strict-boundaries/Administrative agents can reduce coordination work, but clinical meaning and patient access can enter the workflow sooner than the label suggests.Wed, 01 Apr 2026 07:30:00 GMTGaurav TiwariEvaluation & assuranceGovernance & privacyAgent authorityHuman oversightIndustry applicationsPrivacy engineeringAgents in practicehttps://softlander.co/insights/blog/turn-ai-governance-principles-into-controls/https://softlander.co/insights/blog/turn-ai-governance-principles-into-controls/A principle becomes operational only when it is translated into requirements, control owners, enforcement points, evidence and responses to failure.Tue, 31 Mar 2026 07:30:00 GMTGaurav TiwariEvaluation & assuranceGovernance & privacyEvaluation evidenceGovernance controlsHuman oversightGovernable agentshttps://softlander.co/insights/blog/embedded-security-when-you-cannot-patch-quickly/https://softlander.co/insights/blog/embedded-security-when-you-cannot-patch-quickly/A device in the field may be unreachable for months and in service for a decade. That constraint should change the architecture, not just the update policy.Mon, 30 Mar 2026 07:30:00 GMTGaurav TiwariSecurity engineeringEmbedded securityThreat modellingSecure by designhttps://softlander.co/insights/blog/security-requirements-the-team-can-test/https://softlander.co/insights/blog/security-requirements-the-team-can-test/"The system shall be secure against unauthorised access" cannot be implemented, reviewed or verified. Security requirements become useful when each one names an actor, a condition and an observable result.Fri, 27 Mar 2026 10:00:00 GMTGaurav TiwariSecurity engineeringRelease assuranceSecure developmentSecure by designhttps://softlander.co/insights/blog/adoption-is-part-of-agent-reliability/https://softlander.co/insights/blog/adoption-is-part-of-agent-reliability/An agent is not reliable if people avoid it, over-trust it or route around its controls. Adoption evidence belongs beside technical performance evidence.Thu, 26 Mar 2026 09:30:00 GMTGaurav TiwariImpact measurementStrategy & transformationAgent reliabilityHuman oversightOrganisational changeAgentic transformationhttps://softlander.co/insights/blog/prompt-injection-is-a-workflow-security-problem/https://softlander.co/insights/blog/prompt-injection-is-a-workflow-security-problem/Prompt injection becomes consequential when untrusted content can steer a model towards privileged tools. Reduce the blast radius by redesigning the workflow around trust and authority.Wed, 25 Mar 2026 08:30:00 GMTGaurav TiwariAgent securityArchitecture & engineeringIdentity and accessPrompt injectionTool securitySecure agentshttps://softlander.co/insights/blog/foundation-model-one-dependency-agentic-system/https://softlander.co/insights/blog/foundation-model-one-dependency-agentic-system/Model selection matters, but production behaviour emerges from an entire dependency graph. Review the system that turns model output into business effects.Tue, 24 Mar 2026 13:00:00 GMTGaurav TiwariArchitecture & engineeringEvaluation & assuranceAgent architectureEvaluation evidenceObservabilityFrontier agentshttps://softlander.co/insights/blog/financial-services-agents-require-more-than-model-testing/https://softlander.co/insights/blog/financial-services-agents-require-more-than-model-testing/Financial agents combine models, data, tools, people and third parties, so assurance has to follow the complete regulated workflow.Mon, 23 Mar 2026 13:00:00 GMTGaurav TiwariAgent securityEvaluation & assuranceGovernance & privacyContinuous assuranceGovernance controlsIncident readinessIndustry applicationsAgents in practice