o latent Family Source

Skip to content

kernel ready 3 cells

cells

Evaluating LLM outputs without fooling yourself
MCP: giving language models real tools
RAG pipelines that actually work in production

all 3 cells →

namespaces

Evals 1
Retrieval 1
Systems 1

navigate

series
tags
archive
about

Index / vocabulary

Tags

15 tags across 3 posts.

#Evals 1
#LLM-as-Judge 1
#Metrics 1
#Testing 1
#Benchmarks 1
#MCP 1
#Agents 1
#Tools 1
#LLM 1
#Protocol 1
#RAG 1
#Retrieval 1
#Embeddings 1
#Vector Search 1
#Reranking 1

run complete — ship it, then log what broke.

categories/

all
Evals
Retrieval
Systems

tags/

#Agents
#Benchmarks
#Embeddings
#Evals
#LLM
#LLM-as-Judge
#MCP
#Metrics

nav/

series
archive
now
about

Legal, family, and brand

Legal

Privacy
Terms
Refunds
Disclaimer

Family

About
Contact
Sitemap
Security.txt

Connect

Newsletter
RSS
GitHub
Sponsors

A working log on building AI systems that survive production — RAG, agents, and honest evals — by Chirag Singhal.

© 2026 Chirag Singhal · latent · MIT

home Home read Latest series Series find Search menu Menu