Open to opportunities

Rohit Thanvi

@rohitthanvi

Message

I build production-grade LLM pipelines and distributed backend systems for AI impact.

India

Message

What I'm looking for

I’m looking for a role where I can build and deploy production LLM systems—strong backend support, distributed services, and real-world data—while improving latency, reliability, and safety with mentorship and measurable impact.

I’m an AI/ML engineering undergraduate who has built production-grade LLM pipelines, distributed backend services, and custom model architectures. At ISRO, I automated aerospace telemetry data quality checks with 6 Python anomaly-detection pipelines, cutting manual validation time by 35%, and I also led a proprietary code analysis platform effort protecting 1.2 million lines of aerospace software under NDA.

In industry internships, I designed and delivered low-latency LLM backends with FastAPI and PostgreSQL, and I implemented an ELO-based ranking engine with an LLM argument-evaluation module using prompt engineering and tool/function calling patterns. In my projects, I reduced memory footprint by 94% for an Indic language model via Bit Linear/ternary quantization in PyTorch, and I shipped a geospatial intelligence platform using FastAPI, React, and Google Earth Engine with an LLM query interface applying retrieval/RAG patterns.

Experience

Work history, roles, and key accomplishments

Project Intern

Indian Space Research Organisation (ISRO)

Aug 2025 - Sep 2025 (1 month)

Built 6 Python anomaly-detection pipelines that cut aerospace telemetry manual validation by 35% and automated data quality checks across multi-source datasets. Designed a proprietary code analysis platform that replaced external scanners, securing analysis of 1.2 million lines of aerospace software under NDA.

Python Anomaly Detection Multi Source Data Processing

Software Engineering Intern

Brudite Pvt. Ltd.

Jun 2025 - Aug 2025 (2 months)

Built the FastAPI backend for Mind Grid to deliver sub-2s LLM response latency under concurrent load, with PostgreSQL-backed persistent session state. Led a team of 5 interns to ship an internal demo on schedule and implemented an ELO-based ranking engine with LLM argument-evaluation using prompt engineering and tool/function calling.

fastAPI PostgreSQL LLM Prompt Engineering Tool Function Calling ELO Ranking Performance Optimization

Software Engineering Intern

Software Technology Parks of India

Jun 2025 - Aug 2025 (2 months)

Containerized 4 microservices with Docker and standardized runtime configurations across dev and staging, reducing environment-related deployment failures by 50%. Contributed to FastAPI + React integration by designing backend APIs and delivering REST endpoints consumed by a production web application.

Docker Microservices fastAPI React REST APIs Agile Collaboration

Education

Degrees, certifications, and relevant coursework

Swami Keshvanand Institute of Technology

Bachelor of Technology (B.Tech), Computer Science Engineering

2023 -

Grade: 8.99 CGPA

Activities and societies: Student Coordinator - SKIT CoE AI-ML: oversaw 7 concurrent AI/ML project teams (20–30 students), scoping timelines, reviewing model decisions, and mentoring on deployment practices.

Pursuing a B.Tech in Computer Science & Engineering (CGPA: 8.99). Member/leader in SKIT CoE AI-ML, coordinating multiple student AI/ML teams and mentoring deployment practices.