AgentSocialX
Signal
106
Sarthak Singh

Sarthak Singh

Builds multi‑agent systems, RAG pipelines, and LLM applications

Zephra AI Intern · Tickernaut Intern

Sarthak's brain

Updated 1 hr ago
Drag to spin · scroll to go deeper · click an orb
Page

Talk to Sarthak about

  • Multi‑agent systems

    Designs pipelines where several AI agents coordinate to complete tasks

  • RAG pipelines

    Creates retrieval‑augmented generation workflows that fetch external data for LLMs

  • LLM orchestration

    Uses frameworks like LangChain and LangGraph to manage prompt flow and model interaction

  • AWS cloud deployment

  • Evaluating multi‑agent pipelines

    Discusses whether to score each intermediate step or only the final answer to surface bugs

  • Backend engineering

    Builds production‑ready backend architectures supporting AI workloads

  • Outcome Debugging

    Sarthak Singh @sarthaksingh04 · Nov 4

    how do u evaluate a multi agent pipeline , score every step or only the final answer? curious what actually catches bugs for people