Looking for a job

Devarshi Patel

@devarshipatel1

AI/ML engineer at Bytes Technolab building RAG assistants and real-time generative AI systems.

At Bytes Technolab, I developed a real-time AI Draft Writing system with LangGraph and Docker, turning support tickets and product materials into a unified knowledge base for drafting and live-chat assistance.

I also developed a serverless design-generation pipeline using AWS Step Functions, Lambda, and AWS SAM. It turns keyword-based inputs and time ranges into prompts and visual assets through LLM-based generation.

For a wedding experience platform, I built image-embedding and similarity-search pipelines, preference-based recommendations, and a RAG-powered conversational assistant. I used AWS Nova, the Claude API, FastAPI, and MongoDB Atlas Vector Search.

At Ganpat University, I worked on research projects in crowd surveillance, multi-camera person re-identification, and vision foundation models. My work includes published research, including papers in Procedia Computer Science, and projects recognized through research presentations and competitions.

Experience

Work history, roles, and key accomplishments

Current

Associate Software Engineer

Current

Bytes Technolab

Jan 2026 - Present (9 months)

Developed a real-time AI Draft Writing system using LangGraph and Docker, and built an AI-based design generation pipeline using AWS Step Functions and Lambda. Also developed an ML-based N2O emission prediction solution and scalable AI services for a wedding experience platform.

LangGraph Docker AWS Step Functions AWS Lambda AWS SAM Machine Learning fastAPI MongoDB Atlas Vector Search RAG Vision Embeddings Similarity Search Claude APIs MME Nova Lite

Project Coordinator Intern

Maharshi Industries

Jun 2025 - Dec 2025 (6 months)

Developed a computer vision-based inspection system for component detection, OCR, and automated defect analysis. Designed image-processing pipelines for industrial inspection and delivered a real-time vision-based project at LoC, Kashmir.

Computer Vision OCR Image Processing Deep Learning Object Detection

Research Intern

Ganpat University

May 2023 - Aug 2023 (3 months)

Developed a deep learning model for crowd surveillance using YOLOv8, trained on diverse datasets including JHU-CROWD++ and government collaborative data.

Deep Learning YOLOv8 Computer Vision Model Training