Open to opportunities

Harshit Vashisth

@harshitvashisth

Message

AI Engineer and data science intern building secure, multi-agent LLM systems for real-world workflows.

India

Message

What I'm looking for

I’m looking for a role where I can build production-grade AI—multi-agent LLM workflows, RAG, and scalable ML pipelines—while improving reliability, security, and uptime in a fast, collaborative team.

I’m an AI Engineer and data science intern focused on turning cutting-edge LLM and retrieval systems into reliable, production-ready solutions. I enjoy building architectures that move fast—parallel multi-agent pipelines, streaming experiences, and automation that actually saves time.

In clinical and legal workflows, I built an Electron/React/Python clinical AI desktop assistant and delivered secure PDF-to-spec automation converting annotated CRFs into SDTM Excel specifications across 8+ CDISC domains. I also implemented a multi-agent legal AI assistant with OpenAI, Gemini, Azure AI, and RAG, cutting research time by 95% with citation-backed search over 250GB+ of data.

Most recently, I deployed 10+ ML pipelines on Databricks & Azure ML, fine-tuned LLMs and RAG models (LLaMA, DeepSeek R1, Hugging Face, OpenAI), and used MLflow and DVC to improve model accuracy by 12%. I care about scalability and operational excellence too—my systems supported 50K+ users with 99% uptime and handled 1M+ requests/month.

Experience

Work history, roles, and key accomplishments

Data Science Intern

Medtek.ai

Jan 2026 - May 2026 (4 months)

Built Clinyx, a clinical AI desktop assistant using Electron, React, Python, and Claude API for clinical programming workflows. Designed a parallel multi-agent system to automate SDTM/ADaM/TFL code generation and delivered secure PDF-to-spec automation converting annotated CRFs into SDTM Excel across 8+ CDISC domains.

Electron React Python Claude CDISC SDTM ADaM TFL PDF To Spec Automation

AI Innovation Incubator (AIII)

Concept2Action

Aug 2025 - Oct 2025 (2 months)

Implemented a multi-agent legal AI assistant with OpenAI, Gemini, Azure AI, and RAG, reducing research time by 95%. Built citation-backed retrieval pipelines over 250GB+ of legal data and developed a FastAPI/React/LangChain platform supporting 50K+ users with 99% uptime.

OpenAI Gemini)Azure AI Lang Chain Citation Backed Retrieval fastAPI React

AI Engineer

Adjibar Tech

Jun 2024 - Jul 2025 (1 year 1 month)

Deployed 10+ ML pipelines on Databricks and Azure ML with Prometheus and ELK, serving 1M+ requests per month. Fine-tuned LLMs and RAG models (LLaMA, DeepSeek R1, Hugging Face, OpenAI) and improved model accuracy by 12% using MLflow and DVC for experiment tracking.

Databricks Azure ML Prometheus ELK LLM Fine tuning RAG MLFlow DVC Vector Databases

Research and Software Engineering Intern

Defense Research and Development Organization (DRDO)

Jun 2022 - Jul 2022 (1 month)

Engineered a packet sending tool for UDP/TCP network testing and optimized a high-throughput traffic generator sending 829K+ packets per minute using multithreading. Built a Java-based chatbot recognizing 100+ patterns to improve software usability and interaction.

Java Multithreading Chatbot Development Traffic Generation Network Testing

Education

Degrees, certifications, and relevant coursework

Jaypee Institute of Information Technology

Master of Technology, Computer Science and Engineering (AI/ML)

2024 - 2026

Grade: CGPA: 7.8

Pursuing a Master of Technology in Computer Science and Engineering (AI/ML) at Jaypee Institute of Information Technology (CGPA: 7.8). Dissertation on Resource Provisioning in CloudSim for Profit Maximization with SLA Violation Minimization.