const jithu = {
role: "AI Software Engineer @ IU School of Public Health",
education: "MS Computer Science — Indiana University (May 2026)",
location: "Kirkland, WA 🌲",
focus: ["RAG Pipelines", "LLM Integration", "Backend Systems", "AI Agents"],
building: "AI Agent Orchestration Platform (FastAPI · LangChain · Groq · Pinecone · Kubernetes)",
philosophy: "Shipping reliable, well-tested AI features end-to-end — not prototypes."
};I build AI-powered backend systems that go to production. At IU's School of Public Health, I shipped a RAG assistant serving 3,500+ undergraduates, and at Bosch I optimised C++ middleware and CV pipelines for vehicle testing. I care deeply about responsible AI, source grounding, and systems that hold up under real load.
| Project | Stack | Status |
|---|---|---|
| 🎓 RAG Assistant @ IU SPH | TypeScript · Node.js · Vector Search · LLM Embeddings | ✅ Production |
| 🏥 Resolvly — Insurance Denial Analyzer | FastAPI · React · Multi-agent · LLM fallback chain · Docker | ✅ Shipped |
| 🧠 MeetMind AI | OpenAI Whisper · GPT · FastAPI · GitHub Actions | 🏆 Hackathon Winner |
| 🎯 | Achievement |
|---|---|
| 👥 | Shipped production RAG assistant to 3,500+ undergraduates at IU |
| ⚡ | Reduced C++ test execution time by 85% at Bosch with SOLID-pattern backend services |
| 🤖 | Cut vehicle feature testing cost by 40% via CNN + OCR integration |
| 🏅 | Luddy Hackathon Winner — MeetMind AI (2025) |
| 📄 | 2× IEEE publications — Face Recognition (HOG + CNN + GCP) and LDR→HDR reconstruction |
| 🔧 | Contributed to iceoryx distributed IPC middleware (used in automotive embedded systems) |
Languages
AI / ML
Backend & Cloud
Databases
Frontend
| 🏛️ | Degree | Period |
|---|---|---|
| 🎓 | MS Computer Science — Indiana University Bloomington | Aug 2024 – May 2026 |
| 🎓 | BE Computer Science — Kumaraguru College of Technology | Aug 2019 – May 2023 |
Coursework: Applied Algorithms · Distributed Systems · Engineering Cloud Computing · Applied ML · Advanced Database Technologies · Software Engineering
IEEE Publications
- 📄 HOG and Cloud Computing based Face Recognition for Attendance Monitoring — Real-time face recognition at 90% accuracy using HOG + CNN + Google Cloud Firestore
- 📄 Low to High Dynamic Image Reconstruction — GAN/CNN-based LDR→HDR reconstruction, 35%+ improvement across Drago, Reinhard, and Mantiuk tone-mapping
| Project | Description | Stack | Link |
|---|---|---|---|
| 🏆 MeetMind AI | AI meeting assistant — transcription, summaries, action items & sentiment. Luddy Hackathon Winner. | OpenAI Whisper GPT FastAPI GitHub Actions |
→ Repo |
| 🏥 Resolvly | Multi-agent platform decoding insurance denial letters into appeal plans in under 20s, with live CMS/NPPES/eCFR lookups. | FastAPI React LangChain Groq Docker |
→ Repo |
| ⚙️ CV MLOps Pipeline | End-to-end computer vision MLOps — Dockerized model service, GitHub Actions CI/CD, automated image-level regression tests on AWS. | Python Docker AWS GitHub Actions |
→ Repo |
📬 Best way to reach me: jicoim@iu.edu · LinkedIn
Last updated: May 2026 · Actively seeking full-time Software Engineering roles