Open to opportunities

Deep Gautam Kapadia

@deepgautamkapadia

Message

Early-career engineering student building multimodal search and carbon-aware LLM routing systems.

India

Message

What I'm looking for

I’m looking for opportunities where I can build multimodal search and LLM systems that optimize real constraints (energy/carbon), learn fast, and collaborate with engineers and researchers to ship impactful prototypes.

I’m a B.E. student in Electrical and Computer Engineering, and I build applied AI systems that connect performance with real-world constraints. I enjoy turning research ideas into working prototypes with clear, measurable outcomes.

On a CSIR research project, I’m working on designing a more efficient solar inverter system by analyzing inverter architectures and components to improve power conversion efficiency. I like tackling problems where efficiency and design trade-offs matter.

I also developed a multimodal image retrieval system using contrastively trained CLIP embeddings to align images and text in a shared semantic vector space. In parallel, I designed an Eco-LLM inference routing environment for carbon-aware LLM query routing, modeling multi-objective trade-offs between inference accuracy, energy consumption, and grid carbon intensity.

I’m comfortable implementing end-to-end pipelines—from image indexing and semantic search to reinforcement learning environment design (FastAPI, reward calibration, and baseline scripts). I’m active in leadership through executive and organizing roles, and I value collaboration and reliable execution.

Experience

Work history, roles, and key accomplishments

Current

Executive Committee Member

Current

Fine Arts and Photography Society

Sep 2024 - Present (1 year 9 months)

Organized photography and creative events by coordinating logistics and improving participant engagement, and collaborated on workshop and exhibition planning for community outreach.

Event Coordination Logistics Management Workshop Planning Team Coordination

Current

CSIR Solar Inverter Research

Current

Thapar Institute of Engineering and Technology

Aug 2024 - Present (1 year 10 months)

Worked on a CSIR research project to design a more efficient solar inverter system by analyzing inverter architectures and components to improve power conversion efficiency.

Inverter Power Conversion System Analysis data preprocessing Python C C++Collaboration Research

TFF Organizing Committee Coordinator

TFF Organizing Committee

Feb 2025 - Apr 2025 (2 months)

Coordinated core festival planning and execution activities, managing event operations, scheduling, and participant experience across multiple teams.

Scheduling Logistics Planning Participant Experience Event Execution Collaboration

Eco-LLM Inference Routing

Thapar Institute of Engineering and Technology

Designed an OpenEnv-compliant reinforcement learning environment for carbon-aware LLM inference routing, modeling trade-offs between accuracy, energy use, and grid carbon intensity.

Reinforcement Learning Carbon Aware Routing fastAPI Hugging Face Python

Semantic Image Search Engine

Thapar Institute of Engineering and Technology

Developed a multimodal image retrieval system using contrastively trained CLIP embeddings to align images and text in a shared semantic vector space for natural-language search.

Semantic Search Python data preprocessing SQL

Education

Degrees, certifications, and relevant coursework

Thapar Institute of Engineering and Technology

Bachelor of Engineering (B.E.), Electrical and Computer Engineering

2024 -

Grade: Current CGPA: 9.60

Activities and societies: CSIR solar inverter research; CLIP-based multimodal image search; Eco-LLM carbon-aware LLM inference routing using an OpenEnv-compliant RL environment (FastAPI/HF Spaces).

Pursuing a B.E. in Electrical and Computer Engineering (current CGPA: 9.60). Working on a CSIR research project and building ML/NLP and retrieval systems (including CLIP-based multimodal search and carbon-aware LLM inference routing).