RVNitesh Kumar
R V Nitesh Kumar — Index 2026AIFull StackDataProduct Engineering
Forward Deployed Engineer · Chennai

Full-stack engineer working across AI agents, RAG systems, enterprise SaaS, geospatial intelligence and product engineering.

Now
Peninsular Research Operation
Shipping
Agents · RAG · SaaS · Geo
Since
2023 · 2.5+ years
See the work 04 FEATURED · 03 EXPERIMENTS

Selected work

Four products I’ve helped ship — from AI agent platforms to satellite analytics. Open any one for the problem, the system and my part in it.

01Data visualisation & governance platformArchitect & full-stack2024 —

DataMocha & Agentic AI Studio

An analytics and governance platform that turns data from many sources into interactive dashboards, tailored KPIs, reports and forecasts — with multi-tier access and task tracking for every unit. I architected its configurable core, made it cloud-agnostic and shipped it 24% lighter.

02AI meeting intelligence & commitmentsFull-stack & AIProduct

Koriva

Meetings end; what was promised usually gets lost. Koriva turns conversations from Zoom, Teams and Meet into minutes, commitments and searchable intelligence.

RNRN · 00:14:32RecZoomTeamsMeet

“Okay — I’ll send the revised proposal by Friday.”

COMMITMENTSend revised proposalOwner RN · due Friday
DECISIONPilot moves to week twoLogged to meeting minutes
ACTION ITEMLoop in finance on costsOwner AK · open
03Client-side document intelligence

AnyDoc LLM

PDF, DOCX and Excel to clean Markdown — parsed by WebAssembly entirely in the browser, so documents never leave the device.

  1. 01UploadPDF · DOCX · XLSX
  2. 02ParseWASM
  3. 03TransformSTRUCTURE
  4. 04PreviewLIVE
  5. 05Export.MD
Angular · WASM · local
Original · PDFWhat the LLM sees · MD
Drag to compare — same page, before and after.
04Satellite intelligence · Government of OdishaGeospatial APIs & maps2024 —
  1. Sentinel-2 imagery
  2. Polygon selection
  3. Spatial query
  4. Vegetation intelligence

FarmBuddy

Satellite imagery turned into region-level crop and land intelligence — draw a polygon, query millions of data points, read the vegetation.

Spectral index
LOW83 CELLS IN POLYGONHIGH$geoWithin: { $geometry: Polygon }
// 2dsphere index

Also built — smaller, sharper

Open source · Tooling · Independent

03 — How I work

I like working where software meets ambiguity.

The best problems rarely arrive as tickets. They arrive as a CFO who can’t trust a report, a meeting where nobody remembers who promised what, a state agency with terabytes of satellite imagery and no way to ask it a question.

My job is to sit close to that problem, find the real requirement underneath, design a system that fits it — then build, deploy and keep tuning it with the people who use it. Frontend, backend, data or model: whichever layer the problem lives in.

  1. 01UnderstandFind the actual problem, not the stated one.
  2. 02DesignShape the system and the data around it.
  3. 03PrototypeGet something real in front of users fast.
  4. 04Integrate & deployWire it into the systems it has to live with.
  5. 05IterateDebug, measure, refine with real feedback.

Trajectory

04 — Experience
2023Web Development
2023 — 24Full Stack
2024 —AI Systems
NOWForward Deployed Engineering
  1. Forward Deployed Engineer

    Taking the systems work closer to the problem — working directly with stakeholders to turn ambiguous requirements into deployed, AI-powered software, then integrating, debugging and iterating in place.

  2. Software Development Engineer

    • —Architected Agentic AI Studio, a multi-tenant agent platform on Azure OpenAI with a RAG pipeline over Pinecone; delivered 5 specialised agents.
    • —Led DataMocha’s move from Azure Blob to MinIO for cloud-agnostic storage, and cut the build size by over 24%.
    • —Built FarmBuddy’s geospatial APIs on MongoDB 2dsphere for millions of Sentinel-2 data points.
  3. Software Development Intern

    Delivered a full-stack e-commerce application — product management, authentication and purchasing — with Angular, ASP.NET Core, MySQL and MongoDB, responsive across desktop and mobile.

  4. Web Developer

    Built web applications with JavaScript, Node.js and Angular — UI polish, backend integration and application optimisation in an Agile team.

  5. B.Tech, Information TechnologyCGPA 8.92

The stack, in context

Tools only matter where they shipped. Hover or focus any technology to see the work it powered.

Frontend
Backend
AI
Data
Cloud
Tools
RAG2 OF 9 PROJECTS
  • DataMocha & Agentic AI Studio
  • Koriva
  • AnyDoc LLM
  • FarmBuddy
  • OpenMetadata
  • PR Intelligence Agent
  • Jelly Notes
  • E-commerce app (intern)
  • Focus Research Labs

Proof

06 — Certifications & recognition
9×Recognition of ExcellencePeninsular Research Operation — across DataMocha, Smart Agri Tech, data governance, UI/UX optimisation, performance engineering, audit compliance and enterprise feature delivery.
PRODUCT IMPROVEMENTKaizen Katalyst AwardFor proposing improvements that sharpened feedback collection and decision-making in DataMocha.
QUALITYDeming AwardFor high-quality, customisable UI experiences that improved usability across enterprise applications.
07 — Contact

Have a difficult problem worth building?