j
jvanku

Jvan T

@jvanku

I build AI systems, web apps and mobile apps that scale

India
English, Hindi
About me
I’m a full-stack AI software developer helping clients build production-ready AI web and mobile apps. - AI chatbots, RAG systems & automation tools - SaaS dashboards, MVPs & internal apps - APIs, auth, databases & cloud deployment - Clean UI flows and scalable architecture - Tech: Python, FastAPI, Next.js, React, Streamlit, MongoDB, MySQL, OpenAI, Gemini, LangChain, vector DBs Expect clear communication, clean code, fast iteration, and practical AI features from idea to launch.... Read more

Skills

j
jvanku
Jvan T
Offline • 
Average response time: 1 hour

See my services

AI Websites & Software
I will do ai app development, ai website, ai chatbot, ai agent, ai integration
From$350 / projectMore details

Portfolio

Work experience

World_Health Organization

Principal AI Engineer

World Health Organization • Full-time

May 2024 - Apr 2026 • 1 yr 11 mos

• Architected large-scale GenAI ETL & retrieval pipelines over unstructured clinical data (EHRs, scanned reports, images) using LLMs, RAG, OCR, classifiers, etc; built relational-aware extraction pipelines transforming unstructured documents into longitudinal OMOP/SDTM datasets, processing 10M+ records while reducing data loss below 5%. • Designed evaluation frameworks (LLM-as-judge, Golden set benchmarking, human feedback loops) to monitor quality, improving extraction/retrieval accuracy by 19% and stabilizing production performance. • Deployed low-latency retrieval query systems for large-scale clinical workloads, supporting global healthcare clients with optimized inference and data ingestion pipelines.

AI Lead

Company • Full-time

Apr 2024 - May 2025 • 1 yr 1 mo

• Led development of agentic multi-objective optimization (MOO) systems for academic scheduling and program management, reducing resource under-utilization by 42% across large-scale institutions. • Managed a team of 10+ engineers. Built and deployed LLM-powered systems for automated course generation, quiz synthesis, and AI voice agents for sales, leveraging RAG pipelines, prompt optimization, and domain-specific dataset creation. • Designed evaluation and benchmarking pipelines (LLM-as-judge, human-in-the-loop with educators, task-specific metrics) to ensure quality, consistency, and reliability before deployment.

Co-Founder & CTO

SaaS • Full-time

Jan 2023 - May 2024 • 1 yr 4 mos

• Architected a multi-agent AI platform with AutoGen & Langchain leveraging LLMs, RAG,CodeContext and vector search (FAISS) over codebases, documentation, and developer context, achieving 85%+ task relevance accuracy in complex engineering queries. • Built a multi-agent orchestration system (planning, retrieval, code reasoning) to decompose and execute large-scale engineering tasks (e.g. migrate a 10+ yr Spring Boot to Node), reducing planning/debugging time by 40%. • Developed multi-objective optimization (MOO) statistical agents for task allocation across teams (skill, availability, dependencies), improving execution efficiency by 30% and enabling an “AI Engineering Manager” for automated roadmap & execution guidance.