Open to opportunities

Prasham Mehta

@prashammehta1

Message

AI/ML engineer building production-ready ML for drug discovery, chip design, and healthcare.

India

Message

What I'm looking for

I’m looking to build end-to-end AI/ML systems on AWS—shipping reliable ML and RAG/LLM solutions with strong monitoring, governance, and real-world integration through REST APIs.

I’m an AI/ML Engineer with 5+ years of experience designing and deploying machine learning solutions across drug discovery, chip design, and healthcare domains. I focus on taking models from data preprocessing and feature engineering through to reliable deployment.

In end-to-end ML development, I’ve built supervised models (including XGBoost and Random Forest) and used GNNs to model complex molecular and protein-protein interaction networks for drug-target efficacy. I also design ML pipelines that integrate with existing engineering workflows.

I bring strong Generative AI execution, including RAG-based research assistants built with LlamaIndex and vector databases, LLM fine-tuning for biomedical retrieval and generation, and AI agents approaches using LangChain. I improve quality using techniques like strict JSON-schema prompting, chunking, and re-ranking.

For productionization, I deploy on AWS (EC2, SageMaker) and expose inference via REST APIs, with monitoring and validation to track drift, latency, and prediction quality. I’m motivated by shipping measurable impact—faster cycles, stronger convergence, and more dependable outputs.

Experience

Work history, roles, and key accomplishments

Software Developer – AI/ML

GenChip AI Pvt. Ltd.

Mar 2026 - Jul 2026 (4 months)

Designed and deployed supervised ML models (XGBoost, Random Forest) to predict VLSI chip performance metrics (timing, power, area), reducing manual design iteration cycles by 30%. Deployed inference endpoints on AWS EC2 via REST APIs and automated drift-triggered retraining with model monitoring and structured logging.

XGBoost Random Forest Python scikit learn Feature Extraction Model Training AWS EC2 AWS REST APIs Model Monitoring

Machine Learning & AI Engineer

SAS-AI

Oct 2025 - Jan 2026 (3 months)

Built a PyTorch Graph Neural Network (GNN) for lung cancer drug discovery and used a RAG-based generative AI research assistant to fine-tune retrieval and generation over 50,000+ biomedical documents. Implemented preprocessing with Hugging Face Transformers and OpenAI APIs, and improved answer relevance with JSON-schema prompting, chunking, and re-ranking; also identified 100+ novel drug candidate

PyTorch RAG LlamaIndex Vector Databases LLM Fine tuning Hugging Face Transformers OpenAI API Prompt Engineering data preprocessing

Data Scientist & Web/Apps Dev

American Academy of Infection Associated Chronic Conditions

Aug 2021 - Aug 2025 (4 years)

Developed a healthcare cost predictor using SQL and real-time processing on 50,000+ patient records, using prompt engineering with OpenAI APIs to generate actionable clinical insights. Created a nonprofit healthcare website serving 15,000+ monthly visitors, collaborating with 12 doctors and increasing awareness metrics by 35%.

SQL Real time Data Processing Prompt Engineering OpenAI API Data Analysis Healthcare Analytics Web Development

Education

Degrees, certifications, and relevant coursework

Syracuse University

Master of Computer Science, Computer Science

2023 - 2025

Activities and societies: Coursework: Assured Programming using Formal Methods; Design and Analysis of Algorithms; Artificial Intelligence.

Completed a Master of Computer Science with coursework in Assured Programming using Formal Methods, Design and Analysis of Algorithms, and Artificial Intelligence.

Atharva College of Engineering

Bachelor of Engineering in Information Technology, Information Technology

2019 - 2023

Activities and societies: Coursework: Data Structures and Algorithms; Operating Systems; Database Management Systems.

Earned a Bachelor of Engineering in Information Technology with coursework covering data structures, operating systems, and database management systems.