Currently ML Engineer at Zynix.ai
Applied AI Researcher & Machine Learning Engineer
I do AI research and ship it in production. First-author papers at CVPR 2026 and ACL 2026, a national IndiaAI Hackathon win, predictive models serving 1M+ patients, and real-time voice pipelines handling 1,000+ concurrent calls.
Peer-reviewed publications and national recognition.
From research prototypes to systems that run at scale.
Machine Learning Engineer
Machine Learning Engineer Intern
Data Engineering Intern
Integrated M.Tech in Applied Geophysics · Specialization in AI/ML
Systems built and shipped independently.
Generative video
Multilingual, temporally consistent dubbing without paid GPU infrastructure, using distributed local speech synthesis with Kaggle T4 inference and VRAM-aware preprocessing. Stable within 16 GB VRAM while preserving background and mouth-region consistency.
Document intelligence
Private analysis of technical PDFs, scans, and tables without third-party APIs. Optimized Qwen2-7B for local reasoning with layout pipelines built on Camelot, img2table, and OpenCV, improving complex table extraction from below 50% to 85%+ within ~14 GB VRAM.
Agentic systems
A decoupled MCP client-server architecture with asynchronous tools that coordinates research, drafting, and image sourcing, automating the workflow from an initial topic to a structured, illustrated PDF.
Computer vision
An OCR and layout-analysis pipeline that converts scanned invoices into structured XML while retaining layout relationships, reaching 87% key-field precision against an 80% cloud baseline.
Depth across the modern ML stack.
For research collaborations, roles, or technical discussions.