Ashish Kaul
Always curious about enterprise AI, agentic systems, and the long arc of CRM - currently building the future of how we sell
Nineteen years in enterprise technology, mostly in the unglamorous middle distance between what gets demoed and what actually ships. This site is where I write longer-form, watch the field, and think out loud.
Three threads, one field.
The topics this site keeps returning to, picked because they're each going through a real, structural change, and because there's still a lot worth saying that nobody is saying clearly.
GenAI, in practice
Where the field is loud, where it's quiet, and which patterns are actually surviving contact with production.
Agentic systems
Multi-agent architectures, MCP and A2A, evaluation harnesses, the unsexy plumbing under the demos.
Enterprise CRM, the long arc
Notes from two decades of large platform transitions, what changes, what doesn't, and what AI is actually doing to the category.
Recent writing
The AI Reckoning: Why Your Compute Bill is Out of Control and How to Fix It
Exploring why enterprise AI costs are spiraling out of control and how organizations can implement the governance structures needed to align spending with actual business outcomes.
Agentic AI Governance: From Policies on Slides to Code in Production
The 2026 Gartner Hype Cycle named agentic AI governance, security, and FinOps as new disciplines. The technology that produced that need is also producing the ways it actually gets enforced.
Embodied AI: Gemini Robotics and the Closing Sim-to-Real Gap
Gemini Robotics ER-1.6 is the first model that treats physical-world reasoning as a first-class objective. What it gets right, what it still doesn't, and why this is the year robotics stopped being its own island.
Formal Math Reasoning: From Silver to Gold at the IMO
A year after AlphaProof reached silver-medal performance at the International Mathematical Olympiad, Gemini Deep Think hit gold — and produced rigorous proofs in natural language, end to end.
Agentic Coding: When the IDE Becomes the Junior Engineer
Claude Opus 4.7 hit 64.3% on SWE-bench Pro — not as a benchmark stunt, but as the floor for what agent-driven development now looks like in real codebases.
Multi-Agent Orchestration Reaches the Production Tipping Point
Why 2026 is the year teams of specialized agents replaced monolithic super-agents, and what enterprises learned the hard way about coordination, cost, and accountability.