Keyurkumar Panchal — AI Engineer

~/about

About

AI Engineer focused on Reinforcement Learning, Computer Vision, and LLM-powered agents. I enjoy working at the edge of research and engineering — building systems that learn from messy, real-world data and turning them into something that actually runs in production. I've shipped Python backends, ERP modules, and full-stack web apps, and I'm always looking for harder problems to solve.

01 Published Research Paper

04 Focus — RL · CV · DL · LLM

02 Live Deployed Apps

~/skills

Skills

Computer Vision

OpenCV Ultralytics YOLO Models DeepSort MediaPipe Image Segmentation Object Tracking Optical Flow Pose Estimation Video Analytics

Deep Learning

PyTorch TensorFlow Keras ONNX CNNs RNNs / LSTMs Transformers GANs Autoencoders Transfer Learning Model Quantization TensorRT

Machine Learning

Scikit-learn XGBoost LightGBM Pandas NumPy Matplotlib / Seaborn Feature Engineering Hyperparameter Tuning Data Preprocessing

LLM, Agents & RAG

LangChain LangGraph RAG Pipelines Hugging Face Transformers ChromaDB FAISS Prompt Engineering Fine-Tuning (LoRA / QLoRA) Embedding Models n8n CrewAI

Reinforcement Learning

Stable-Baselines3 Gymnasium Ray RLlib OpenAI Gym PettingZoo TorchRL TensorBoard

Languages & Scripting

Python SQL HTML CSS XML YAML

Backend, Tools & MLOps

Django Flask FastAPI Git Docker MLflow Weights & Biases Power BI Linux / SSH PostgreSQL REST APIs

~/experience

Experience

Apr 2024 — Jul 2024

Python Developer — Warlock Technologies

Gandhinagar, Gujarat

Designed and developed custom Odoo modules tailored to client needs, enhancing ERP capabilities.
Integrated third-party applications and ERP systems to improve workflow automation and data consistency.
Provided technical support, debugging, and performance optimization for deployed Odoo solutions.
Collaborated with cross-functional teams to translate requirements into scalable software solutions.

Jan 2023 — Jun 2023

Python Developer Intern — Axisray Technologies

Ahmedabad, Gujarat

Built and maintained Django-based web applications, focused on back-end functionality and database management.
Developed RESTful APIs for data exchange between front-end interfaces and server-side logic.
Participated in code reviews and agile development cycles, strengthening teamwork and software quality.
Improved application performance by optimizing queries and refining application structure.

Jun 2022 — Jul 2022

Data Science Intern — Vnurture Technologies

Analyzed datasets using Python to extract insights, visualized using Power BI.
Performed data cleaning, transformation, and modeling to derive business intelligence.
Completed a capstone project on Employee Salary Analysis, identifying trends and presenting results.

~/projects

Pinned Projects

Three of the more involved systems I've built, each with its own deep dive — architecture, methodology, and real results.

Pinned Deep RL · Quant Finance

Autonomous Options Trading & Hedging with Deep RL

A MaskablePPO agent that trades and dynamically hedges NIFTY50 weekly straddles from a 69-feature option-Greeks state space — published at GCon 2026.

93.75%Win Rate

47.66Profit Factor

README.md → Pinned LangGraph · LLM Agents

VocalSQL — Natural Language to SQL

A self-correcting 7-node LangGraph agent that turns plain English into safe SQL, backed by RAG schema retrieval and a 9-layer anti-hallucination pipeline.

7Pipeline Nodes

9Guardrail Layers

README.md → Pinned LLM · FastAPI