Hey, Nikhil here. Engineer at heart, building the data layer of AGI. I’m big on collaboration, staying curious, and building meaningful solutions that actually solve problems.
Whe... Show more
nik-13
@nik-13
Nikhil Pareek
Building the data layer of AGI @FutureAGI | Founder & CEO | Forbes Tech Council
About
Contributions
Articles
- The four axes generic RAG eval forgets for search agentsAug 18, 2026
- How to debug a multi-agent system by walking the span treeAug 14, 2026
- Your custom metric is scoring noise unless it has these three partsAug 10, 2026
- # Evaluating Pydantic AI agents that call MCP toolsAug 08, 2026
- # Evaluating LLM tool use is a four-step contract, so score four numbersAug 06, 2026
Comments
- @[Josaphatstar] Sure. DSPy is the easiest place to start, and TextGrad...Aug 06, 2026
- Love that you benchmarked inside your real pipeline, not a public lead...Jul 15, 2026
- The loop is the easy part' is exactly right. The one I'd push further ...Jul 15, 2026
- @[juhist] Honestly, yeah, it's context-dependent. The metrics fall out...Jul 15, 2026
- Interesting take. Would be curious whether the folder structure measur...Jul 02, 2026
chevron_left