Merit Is Not Self-Executing
Consequential outcomes are shaped by more than competence and rules. Why good work is not self-executing, and what capable people miss.
StrategySeries
July 19, 2026 · 11 min readReadEssays and field manuals on enterprise AI, secure systems, architecture, data centres, and the statecraft of navigating power and influence in the rooms where consequential decisions get made.
Series
The Enterprise AI Architect's Handbook
NewThe complete working knowledge of the role: twenty-nine domains from operating model and retrieval to agents, evaluation, security, platform architecture and production readiness.
13 essaysStart reading
Enterprise AI Strategy
NewEssays on adopting AI as an operating-model change: ownership, pilot-to-production gates, governance, operations, incidents and cost per useful outcome.
9 essaysStart reading
Designing Secure AI Agents
Essays on deploying AI agents without surrendering security, permissions, auditability or human control.
5 essaysStart reading
Designing Scalable Microservices
Essays on service boundaries, contracts, data ownership and the failure modes that decide whether a system holds in production.
5 essaysStart reading
Data Center Project Management in India
Essays on building data centres in India: site, power, cooling, approvals, procurement, commissioning and handover.
10 essaysStart reading
Statecraft
Essays on navigating power dynamics, negotiating, and establishing influence with the people who can impact the outcome you want.
2 essaysStart reading
Selected Writing
AI Agent Prompt Injection Risk Scorecard: How to Assess Agents Before Production
Score AI agent prompt injection risk: source exposure, data sensitivity, tool access, approval gates, and runtime controls. A 12-dimension enterprise assessment.
AICybersecurity
July 19, 2026 · 27 min readReadTenant Context in Multi-Tenant Microservices: How to Design Safe Tenant Boundaries
Tenant context in microservices is a security boundary — how to derive, validate, enforce, and audit it across APIs, queries, jobs, webhooks, and audit logs.
TechnologySeries
July 19, 2026 · 19 min readReadData Center Approvals in India: A Project Manager's Checklist for Land, Power, Fire, Building, Environment, and Telecom
How to manage data center approvals in India: land, building, fire NOC, power, DG, water, environment, telecom, and a tracker template.
Data CentersTechnology
June 13, 2026 · 19 min readReadLatest Essays
Agent Orchestration: Frameworks, Durable Execution and Framework-Independent Design
LangChain, LangGraph and AutoGen as runtimes, not architectures: checkpoints, HITL, durable execution, versioned playbooks and lock-in.
AISeries
September 16, 2026 · 30 min readReadFine-Tuning, RAG or Prompting: How an Architect Decides
Most gaps teams try to train away are instruction or knowledge problems. When fine-tuning is justified, and how to govern data and rollout.
AISeries
September 16, 2026 · 36 min readReadMCP Architecture and the Enterprise Tool Gateway
Tool calling is a privileged execution boundary: tool design, permissions, credential brokering, MCP architecture and the tool gateway.
AISeries
September 16, 2026 · 39 min readReadModel Strategy: Selection, Gateways, Routing and Fallbacks
Models as replaceable compute: selection criteria, gateways, registries, complexity routing, fallbacks, cascades, quotas and self-hosting.
AISeries
September 16, 2026 · 39 min readReadMulti-Agent Systems: When They Help, and How They Fail
When a second agent is justified, the coordination patterns that work, and the distributed-systems failures a multi-agent design inherits.
AISeries
September 16, 2026 · 33 min readReadAgent Architecture: Loops, Planning, Verification and Termination
An enterprise agent is a bounded decision-and-action system. Its anatomy, control loops, durability, verification, authorization and termination conditions.
AISeries
August 16, 2026 · 36 min readReadAgent State and Memory Architecture: Scoping, Retention and Provenance
State, memory and model context are three different things. How to persist, scope, retain, retrieve and govern each one without poisoning or leaking data.
AISeries
August 16, 2026 · 30 min readReadTools from the Work
Reusable templates and operating documents taken from real projects.
AakashX is written by Aakash Ahuja. About the author