Looking for a job

Rishet Mehra

@rishetmehra

Message

AI and full-stack developer building LLM pipelines and agent systems.

India

Message

What I'm looking for

I’m looking to build production-ready LLM and CV agent systems, ship reliable web experiences, and keep raising data quality through rigorous evaluation, iteration, and human-in-the-loop safety.

I’m an AI Research & Development intern focused on turning messy, real-world data into reliable intelligence. At Zanista AI (UK), I built web crawlers for JS-rendered pages, paywalls, and deduplication—ingesting 10,000+ financial news articles daily—and shipped LLM pipelines for summarization, NER, sentiment, and topic clustering to cut manual review time and improve downstream analytics data quality.

I also bring strong product instincts from UI/UX & Web Development: at BharatFare (UK) I triaged a 281-issue backlog and shipped fixes for 40+ React / Next.js bugs to stabilize user flows and reduce load times, while translating competitor findings into roadmap-adopted UX improvements. Through projects like SpaceATC (multi-agent safety coordination and onboard 6-DOF vision), Codex-spend (an npm CLI for token-burn analysis), VidWise (RAG + evaluation dashboards), and SkillGraph (Node2Vec + GNN + LightGBM ranking), I’m driven by rigorous evaluation, human-in-the-loop safety, and shipping systems that actually perform.

Experience

Work history, roles, and key accomplishments

UI/UX & Web Development Intern

BharatFare Travel & Technologies (UK)

Nov 2025 - Jan 2026 (2 months)

Triaged a 281-issue backlog and shipped fixes for 40+ React/Next.js bugs, cutting page load times and stabilizing core user flows. Conducted competitor analysis that identified 3 critical UX gaps and ensured all fixes were adopted into the product roadmap.

React Next.js UI UX Triage Performance Optimization Competitor Analysis Product Roadmapping

AI Research & Development Intern

Zanista AI (UK)

Jul 2025 - Sep 2025 (2 months)

Built web crawlers for JS-rendered pages, paywalls, and deduplication to ingest 10,000+ financial news articles daily with zero manual cleanup. Shipped LLM pipelines for summarization, named-entity recognition, sentiment scoring, and topic clustering to reduce manual review work and improve downstream data quality for financial analytics.

And Deduplication Named Entity Recognition Sentiment Analysis Clustering Financial NLP Summarization