# Poonam Patil

Poonam Patil is a remote worker on Himalayas, Lost & Found – Object Detection & Retrieval System at Self, based in India.

Source: https://himalayas.app/@poonampatil2

```text
Machine learning research intern at IIT Bhubaneswar who built an adaptive federated LLM fine-tuning framework.
```

| Detail | Value |
| --- | --- |
| Availability | Looking for a job |
| Location | India |
| Authorized to work in | India |
| Salary expectations | 70k+ INR |
| Job categories | Machine Learning Engineer, Software Engineer, Backend Engineer |
| Skills | Machine Learning, Federated Learning, Deep reinforcement learning, D3QN, LLM Fine tuning, LoRAs, QLoRA, Async Aggregation, RAG, Vector Databases, Similarity Search, Embeddings, Prompt Engineering, Agentic Workflows Design, Natural Language Processing |
| LinkedIn | https://linkedin.com/in/poonam-patil-a72875242 |

## About

```text
At IIT Bhubaneswar, I built ARA-FL, an asynchronous federated LLM fine-tuning framework that uses a Double DQN agent to select LoRA rank and QLoRA bit-width. It cut average VRAM by up to 70% and cross-entropy loss by 27% compared with FedAvg.

I also built and deployed a legal-document RAG chatbot using FAISS, Streamlit, Anthropic Claude, and OpenAI. In a separate object detection and retrieval project, I created a CPU-only pipeline using YOLOv11s, BoT-SORT, CLIP, and FAISS to find objects’ last-seen times and frames.
```

## Experience

### Lost & Found – Object Detection & Retrieval System

Self. May 2026 - Aug 2026

```text
Architected a 5-stage pipeline for object detection and retrieval using YOLOv11s, BoT-SORT, CLIP, and FAISS, processing live camera feeds at 6-8 FPS on CPU. Designed a dual-store backend with SQLite and FAISS for metadata and embeddings.
```

Skills: Python, YOLOv11, BoT-SORT, CLIP, Faiss, SQLite, Streamlit, Computer-Vision, Object-Detection, Object-Tracking

### Research Intern (Machine Learning)

IIT Bhubaneswar. Jan 2024 - Jan 2025

```text
Developed ARA-FL, an asynchronous federated LLM fine-tuning framework using Deep RL, achieving up to 70% VRAM reduction and 27% CE loss improvement. Engineered training systems including staleness-weighted aggregation and energy measurement.
```

Skills: Python, PyTorch, Federated-Learning, Deep-reinforcement-learning, D3QN, LoRAs, QLoRA, FedAvg, FedProx, Scaffolding, FedASK, Perplexity, ROUGE-L, BERTScore, Exact-Match

### LegalAssistant Chatbot

Self. Jan 2024 - Feb 2024

```text
Built and deployed an LLM-powered RAG application end-to-end, ingesting legal documents into a FAISS vector store and serving answers via Streamlit. Integrated Anthropic Claude and OpenAI APIs, grounding responses in retrieved sources to limit hallucination.
```

Skills: Python, RAG, Faiss, Streamlit, Anthropic-Claude-API, OpenAI-API, Prompt-Engineering, Vector-Databases

## Education

### Indian Institute of Technology Bhubaneswar

Bachelor of Science, Electrical Engineering. 2023 - Present. Grade: 7.83

```text
Pursuing a Bachelor of Science in Electrical Engineering with a GPA of 7.83.
```

## Tech stack

- GitHub
- GitHub Actions
- NumPy
- Pandas
- PostgreSQL
- SQLite
- Node.js
- Django
- Tailwind CSS
- OpenCV
- JavaScript
- Python
- Java
- JSON
- C++
- PyTorch
- scikit-learn
- Streamlit
- NLTK
- Git
- SQL
- XGBoost
- LangChain
- OpenAI API
- Anthropic Claude API
- Faiss
- Matplotlib
- REST APIs
- HTML
- CSS

To contact Poonam, message them from this profile with a Himalayas employer account: [Hire on Himalayas](https://himalayas.app/hire).

## People also viewed

- [Sai Vishwas](https://himalayas.app/@saivishwas): AI Intern, Sree Tech Tammina Software Solutions Pvt. Ltd.
- [Farhan Ahamed User](https://himalayas.app/@farhan): Research Intern, ArchLabs (RAPID), AIML Department, PES University
- [ANUSHKA AGARWAL](https://himalayas.app/@anushkaagarwal): AI/ML Summer Intern, Zenithive
- [Ujjwal Pandey](https://himalayas.app/@iamujjwalpandey): AI Research Intern, Indian Institute of Technology, Bhubaneswar
- [Gokul Venkatareddy](https://himalayas.app/@gokulvenkatareddy): ML Intern, Omilos
- [Siddharth Upl](https://himalayas.app/@siddharthupl): Chair, IEEE Computer Society Student Chapter, Mahindra University
- [Centauri 1310](https://himalayas.app/@centauri1310): Undergraduate Student Researcher, Indian Institute of Technology, Jodhpur
- [Akash Kar](https://himalayas.app/@akashkar): Health Systems Research Intern, Shalink & Reyx Infinity Research Hub Pvt. Ltd.
- [Rohit Kumar](https://himalayas.app/@rohitkumar104): AI Research Intern, Indian Institute of Technology (IIT) Bhubaneswar
- [Poonam Patil](https://himalayas.app/@poonampatil1): Research Intern (Machine Learning), IIT Bhubaneswar
