Skip to content
Vidit-labPublic

About

No description, website, or topics provided.

Resources

Stars

0 stars

Watchers

0 watching

Forks

Latest commit

Β 

History

3 Commits

Folders and files

NameName
Last commit message
Last commit date
Β 
Β 

Repository files navigation

Typing SVG

πŸ“ Udaipur, IndiaΒ Β β€’Β Β πŸŽ“ IIT Bhilai β€” B.Tech. Data Science & AI, 2028Β Β β€’Β Β πŸ† Codeforces Specialist


OpenLake Email LinkedIn Codeforces GitHub Medium X

🧠 About Me

engineer:
  name: "Vidit Shrimali"
  role: "AI Engineer & Competitive Programmer"
  focus: ["Machine Learning", "Multi-Agent Systems", "Real-world Deployment"]
  philosophy: "Bottom-up tinkering > Top-down theories"
  currently: "B.Tech in Data Science & AI @ IIT Bhilai (CGPA: 8.97/10)"

I build AI systems that leave the lab and actually ship β€” from a computer-vision pipeline deployed across 54 government health centres, to multi-agent voice systems making real outbound calls, to autonomous trading agents running 24/7. I compete on Codeforces to stay sharp on algorithms, and apply the same rigor to production ML.

πŸ’Ό Experience

Machine Learning Engineer Intern β€” Anveo Innovations

Aug 2025 – Dec 2025 Β· Chhattisgarh, India

Built and deployed an AI-powered child growth assessment system for the District Collector of Dhamtari β€” automated height & weight estimation directly from images, replacing manual anthropometric measurement.

  • Collected and curated data from 8,568 children across 258 centres, deployed across 54 centres
  • Built an end-to-end inference pipeline (body segmentation β†’ regression) served via AWS for scalable field use

Children Data Centres Deployed Centres Surveyed MAE Weight MAE Height R2 Weight R2 Height

πŸš€ Featured Projects

Analyzed 100,000+ Codeforces records to uncover the real drivers of rating progression, then turned the findings into an interactive analytics platform.

  • Feature-engineered log-linear regression models on contest activity & consistency metrics
  • 10 interactive visualizations delivering data-driven improvement insights

Records R2 Score Visualizations

A Hindi-first multi-agent voice system that autonomously qualifies leads and syncs to CRM in real time using the Gemini Live APIs.

  • Production-grade agent orchestration on GCP with real-time tool use and observability
  • Validated live in the field, not just in simulation

Concurrent Agents Latency Live Calls Cost

A self-improving, 24/7 crypto trading platform (ccxt/Binance) orchestrating 8 LLM agents across a LangGraph pipeline with code-enforced risk guardrails.

  • Retrieval-augmented memory layer on self-hosted Supermemory β€” 5 memory types, 768-dim ONNX vectors, hybrid semantic search
  • Full-stack React dashboard backed by a 20-endpoint REST API

LLM Agents Cycle Time Memory Types Tests Endpoints

πŸ› οΈ Tech Stack



LangChain LangGraph RAG Multi Agent Systems LLM Applications

πŸ“« Let's Connect

Email LinkedIn Codeforces Medium X



Footer typing

About

No description, website, or topics provided.

Resources

Stars

0 stars

Watchers

0 watching

Forks

Releases

Packages

Contributors