Headline

4x Hackathon Winner

About

I build AI-native products for complex, high-stakes environments where messy human habits meet probabilistic models. Where software has to work reliably under real-world constraints. I hold a B.Tech (Hons.) in Computer Science & Engineering (Data Science) with a Minor in Psychology. This combination shapes how I approach product. I start with user friction and cognitive ergonomics before evaluating technical architecture. Rather than treating AI as an autonomous black box, my focus is on human-in-the-loop systems, deterministic guardrails, and intuitive interfaces that earn user trust. I let user behavior drive feature decisions, willingly shelving over-engineered ideas when simpler, high-utility flows drive better task comprehension. I advocate for the design of more effective AI interactions, ensuring that the tool being created, is truly optimised for the people who use them. Beyond technical execution, I am a firm believer that the best solutions emerge from diverse viewpoints. I am equally at ease taking the lead on complex, independent research as I am working within a multidisciplinary team to meet tight deadlines. My background in Psychology has helped me navigate team dynamics, foster clear communication, and resolve conflicts, ensuring that a project's technical goals remain aligned with the collective vision and the end-user's needs. I am passionate about using data to solve social problems from a human-centered perspective. I'm also an avid watcher of cricket and I love playing video games with my friends in my free time!

Location

Bengaluru

AI stack

antigravity

claude code

claude

codex

cursor

gemini

Work experience

Data Science Intern at Amagi, Jun 2026 - Present · 4 mo, Bengaluru, • Crafted active learning pipelines for computer vision models, prioritizing low-confidence video frames to cut manual annotation volume by ∼30% and accelerate model iteration velocity. • Orchestrated data annotation operations by establishing standardized acceptance criteria, review queues, and daily feedback loops with annotators, resolving class ambiguities to boost labeling throughput by 25%. • Targeted real-world model failure modes by building heuristic sourcing pipelines, curating high-signal public video data to fortify model robustness against rare visual edge cases.

AI & Data Intern at Prognos Labs, Jan 2026 - Jun 2026 · 5 mo, Remote, • Spearheaded the product architecture for a clinical documentation tool for Indian OPDs; drove the pivot from custom models to multimodal LLMs to deliver draft notes in <56 seconds, cutting documentation time by 60%. • Suppressed hallucinations to 5% and achieved 99.5% speaker diarization accuracy. • Diagnosed a 46% clinical detail drop-off during summarization; restructured the response schema to boost Medical Term Recall by ~27%. • Designed clinician-centric interaction flows, including an inline “Magic Input” AI co-pilot, visual diff tracking (additions/deletions), version rollbacks, and interactive onboarding spotlights to maximize physician trust. • Architected multi-tenant subscription tiers with automated trial provisioning and longitudinal patient profile memory, enabling cross-encounter continuity across 10 major Indian languages without semantic bleeding.

Summer Intern at Hitachi Energy, Jun 2024 - Aug 2024 · 2 mo, Bengaluru, • Led a 3-person intern team that conceptualized and deployed an internal knowledge chatbot via MS Power Virtual Agent, streamlining new-hire discovery journeys and cutting recurring HR/IT inquiry tickets. • Mapped enterprise data discovery friction and architected a centralized Power Apps portal for internal T-code management, eliminating data lookup bottlenecks.

Education

B.Tech. (Hons.) Computer Science and Engineering – (Data Science), Specialization in Deep Learning, Minor in Psychology, Vidyashilp University

Vidyashilp Academy

Public links

github: https://github.com/vineetnagv