Muthu Bommiah’s working knowledge on governed AI, context engineering and AI-native software delivery, kept current by an LLM librarian; each note says whether he has reviewed it. Every note carries its sources and when it was last verified.
51 notes, 51 reviewed by Muthu so far. Each one shows its review state, when it was last verified and the sources it draws on. The essays themselves live on muthub.org.
Themes
- Context engineering: Trusted, governed knowledge that people and agents can rely on: context layers, knowledge formats, retrieval and provenance (6 notes)
- Governed AI: Guardrails, provenance, identity and policy enforced by design for AI systems and agents, including architecture as code (15 notes)
- AI-native SDLC: How software gets built when agents write much of the code: coding agents, specs, verification and engineering productivity (10 notes)
- Data & AI platforms: Data and ML platforms, lakehouses, MLOps, integration and the infrastructure AI runs on (11 notes)
- Leading AI change: Strategy, operating models, investment cases and leadership of AI change in large regulated organizations (9 notes)
Recently updated
- A data engineering agent in practice
- A2A and MCP: complementary, not competing
- Adopting MCP in a bank: the four risks to solve first
- Agent design is still hard
- Agent-first data architecture: five principles
- AI is an amplifier, not a silver bullet
- AI-Augmented Software Development Lifecycle
- An AI operating model for a bank: ten decisions to make before the twentieth use case