Skip to content
View deep-chokshi's full-sized avatar

Block or report deep-chokshi

Block user

Prevent this user from interacting with your repositories and sending you notifications. Learn more about blocking users.

You must be logged in to block users.

Maximum 250 characters. Please don’t include any personal information such as legal names or email addresses. Markdown is supported. This note will only be visible to you.
Report abuse

Contact GitHub support about this user’s behavior. Learn more about reporting abuse.

Report abuse
deep-chokshi/README.md

Hi πŸ‘‹, I'm Deep Chokshi

AI Engineer | LLM Training & Post-Training | Agentic AI | RAG | GenAI
Building, training, and orchestrating intelligent systems β€” from data and gradients to autonomous agents


πŸ’‘ About Me

  • πŸ§ͺ Working on LLM training & post-training β€” reinforcement learning gyms for tool use, GRPO and policy-optimization pipelines, reward modeling, and SFT
  • πŸ€– Building agentic systems with LangChain & LangGraph β€” multi-step reasoning, tool-calling, and stateful multi-agent orchestration
  • πŸš€ Shipping production-grade GenAI applications with FastAPI, vector stores, and Enterprise-grade RAG architectures
  • 🎀 Hosting webinars & workshops on Prompt Engineering, Agent-based LLM systems, and LLM evaluation

🧠 Skills & Tech Stack


πŸ“¦ Featured Projects

Project Description Tech
πŸ‹οΈ RL Gym for Tool Use Custom RL environment to train LLMs on multi-tool, multi-turn agent tasks β€” reward shaping, trajectory rollouts, and verifier-based scoring NVIDIA NeMo-Gym, PyTorch, vLLM
🎯 GRPO Training Pipeline Group Relative Policy Optimization pipeline for post-training LLMs on tool-use and reasoning traces, with reference-model KL control and reward aggregation GRPO, PyTorch, Hugging Face TRL
πŸ•ΈοΈ LangGraph Multi-Agent Orchestrator Stateful agent graphs with planner / executor / critic loops, tool routing, memory, and human-in-the-loop checkpoints LangGraph, LangChain, FastAPI
🧠 LLM Eval & Reward Modeling Scenario-based evaluation harness with pass@k, tool-call verifiers, and reward-model training for preference data Python, LangSmith, TRL
🧾 Legal Contract AI Reviewer Agentic AI bot that reviews contracts for risks using LangGraph + Azure OpenAI LangChain, RAG, Azure
πŸ€– Job Application Automator Automated ATS with resume parsing, answer generator, and Power BI dashboard FastAPI, PostgreSQL, GPT
πŸ“Š Price Intelligence Bot AI-based price matcher & extractor from competitor eCommerce sites GPT-4o Search, Crawler, Python
🧠 Prompt Engineering Toolkit Set of reusable, tested prompt templates with LangSmith evaluation logs PromptFlow, LangChain, FastAPI

πŸ“‚ Archive β€” Previous Work

πŸ—‚οΈ I recently moved to this GitHub account. My earlier projects (68+ public repos across GenAI, RAG, agentic apps, and more) live on my previous profile β€” feel free to browse them there:

πŸ‘‰ github.com/deepchokshi


πŸ“ˆ GitHub Stats β€” Current Account


🌍 Connect With Me

LinkedIn Email


🧭 β€œGreat AI systems don’t just respond β€” they act, adapt, and evolve.”

Popular repositories Loading

  1. ToolBench ToolBench Public

    Forked from OpenBMB/ToolBench

    [ICLR'24 spotlight] An open platform for training, serving, and evaluating large language model for tool learning.

    Python

  2. nvidia-rl-gym nvidia-rl-gym Public

    Forked from NVIDIA-NeMo/Gym

    Build RL environments for LLM training

    Python

  3. nl-sql-krisha nl-sql-krisha Public

    Python

  4. deep-chokshi deep-chokshi Public

  5. mindmesh mindmesh Public

    One brain. All your AI tools. Fully in sync