I’m Shivam Kumar. I lead cross-functional teams building RLHF/SFT data programs, multimodal evaluation tooling, and operational systems that make machine learning work at production scale.
Technical leader building ML/AI systems and scaling teams. I’ve led programs spanning hundreds of engineers, contributed to multi-million dollar outcomes, and stayed hands-on across LLMs, computer vision, and product engineering.
What I’m best at
I work best where product ambition and execution complexity meet: model evaluation, training-data quality, platform tooling, process design, and the systems that help teams ship reliably at speed.
RLHF/SFT data programs, multimodal evaluation, QA automation, and end-to-end systems that support product-grade ML delivery.
B.Tech (Computer Science), IIT Goa (2020) · CPI: 9.22/10.0
Built and managed teams from small pods to 300+ contributors, with experience across hiring, QA systems, performance tracking, and cross-functional operating rhythm.
Structured, hands-on, and product-oriented. I like systems that are measurable, teams that can move quickly, and tooling that reduces coordination overhead without lowering quality.
Roles where I combined technical depth, systems thinking, and team leadership to produce measurable outcomes.
Led and scaled cross-functional teams of 300+ engineers delivering RLHF/SFT datasets and multimodal evaluation programs for top AI companies.
Progression: Trainer → Team Lead (5) → Project Lead (40) → Delivery Manager / Technical Lead (300+)
Built consumer-facing blockchain products and tooling across Solana and NEAR using Rust and TypeScript.
Applied computer vision and machine learning to production geospatial workflows, shipping systems that improved both coverage and operational efficiency.
Built AI-enabled ed-tech experiences spanning voice models, mobile apps, and backend systems.
Built an ML-based binary vulnerability detection engine using hierarchical filtering and locality-sensitive hashing, achieving more than 90% detection accuracy.
Selected patent and academic highlights.
Co-invented an ML system generating 3D digital surface models from images using computer vision, reducing reliance on LiDAR and enabling operations in regions without dependable elevation data.
Technical Excellence Award (IIT Goa, 2018–2019) · JEE Advanced AIR 3120 (2016)
A few public builds and technical artifacts.
CLI tools and frontend templates to accelerate NFT minting launches on NEAR.
Evaluation tooling to compare multimodal model predictions against ground truth using cloud infrastructure and LLM APIs.
Core technical stack and delivery range.
Agile/Scrum, performance tracking, QA systems, system architecture, and cross-functional team leadership.
The best way to reach me is email. LinkedIn and GitHub are linked below as well.
Start here
If you want to discuss AI delivery at scale, product-grade ML systems, or a collaboration, email is the cleanest way to start the conversation.