Chiranjeev ThomasSenior full-stack engineer building agentic AI

Open to new rolesBengaluru

With 7+ years of experience, I build AI products that real people use, from search that understands plain language to agents that get work done safely, with a human in the loop where it matters.

Chiranjeev Thomas with his Ragdoll cat resting on his shoulder
Psst. Ask me about my human.I’ve napped on his keyboard through every project · here to helpHow I built Miso, and why →
Connecting…
Meow! I’m Miso. I’ve watched Chiranjeev build every one of his projects, so ask me anything about his work. I’ll dig through his portfolio and show you exactly where I found each answer.
Try asking

Projects

AI engineering and full-stack work, plus products I’ve shipped

Showing 6 projects

At Grabbit
  • AI engineering
  • Full-stack

Conversational product search with an LLM

Shoppers describe what they want in plain language; an LLM turns it into an Elasticsearch query. Non-technical users can now run searches that used to need filters and exact keywords.

  • LLM
  • Elasticsearch
  • Spring Boot
  • React Native
  • AWS

Demo: in Grabbit app · Code: private

Live
  • AI engineering

Miso — the agent on this site

Miso is the chat on this site. It searches by meaning and by keyword, answers only from what it found, shows its work in a live trace, and caches answers per deploy so a repeated question costs no LLM call.

  • Next.js 15
  • Groq · gpt-oss-120b
  • Pinecone integrated inference
  • BM25
  • Netlify Blobs
  • Netlify
Live
  • AI engineering

Agent observability & evals for InvoiceChaser

Every InvoiceChaser agent run is traced step by step, and a golden set of 18 disputes grades tool choice, approvals and tone on every prompt or model change. Building it found four real bugs.

  • Python
  • LangChain middleware
  • Supabase
  • LLM-as-judge
  • GitHub Actions
SaaS product
  • Product
  • Full-stack

DevReferral — developer career platform

Helps developers upskill, connect and get referred to roles at top tech companies. Built and ran it as founding engineer — 20+ features, the move to microservices, CI/CD and A/B-tested growth.

  • React
  • Next.js
  • Node.js
  • Express
  • MongoDB
  • TypeScript

Demo: no longer online · Code: private

Experience

  1. Jul 2025 — Present

    Grabbit

    SDE-2 · Early engineering hire

    Last-mile delivery platform · 14,000+ users · 44,000+ orders

    • LLMs
    • Elasticsearch
    • Spring Boot
    • React Native
    • React
    • AWS Lambda
    • DynamoDB
    • Redis Streams
    • AI · LLMShipped LLM conversational search to production — natural language translated into Elasticsearch queries, so non-technical users can run complex product searches and the catalogue is far easier to discover. Case study →
    • B2C → B2BLed the platform’s move from a B2C delivery app to a hybrid B2C + B2B marketplace — across the React Native app, React web portal and Spring Boot + AWS backend.
    • New revenueDesigned and launched the full merchant ecosystem — mobile app, web portal and APIs, with warehouse-scoped data isolation, onboarding independent restaurants and shops across 3 warehouses and unlocking a new B2B revenue stream.
    • −40%Architected real-time GPS order tracking — on RESTful APIs and Redis Streams, giving customers live order status and cutting support inquiries by 40%.
    • 99%+Built an event-driven notification pipeline — (Spring Boot, AWS Lambda, Firebase FCM, DynamoDB) with 99%+ delivery and zero blocking on order creation, eliminating missed merchant orders.
  2. Mar 2019 — Jul 2025

    DevReferral (ReferFee)

    Founding Software Engineer

    Career platform helping developers upskill, connect and get referred to top tech companies

    • React
    • Next.js
    • Node.js
    • TypeScript
    • Microservices
    • CI/CD
    • TDD
    • 0 → 1Owned the full development cycle as the founding engineer — planning, building, debugging, shipping and maintaining, while validating customer problems, running interviews and defining product KPIs.
    • +30% usersGrew the user base 30% — through targeted A/B testing and insights from user interviews.
    • +40%Designed and shipped 20+ end-to-end product features — with UX designers, lifting user engagement 40%.
    • −40%Moved the platform to microservices as it scaled — improving performance and maintainability and cutting deployment times 40%.
    • −60%Cut page load times 60% — in the React and Node.js web apps through code optimisation and asset management.
    • +50%Built and ran the CI/CD pipeline — raising release frequency 50%, and introduced test-driven development across the codebase, reducing bugs 25%.
    • MentorMentored junior engineers — through regular code reviews and knowledge-sharing sessions on engineering best practices.
  3. Jan 2019 — Mar 2019

    dDriven

    Software Engineer Intern · Platform team

    Industry 4.0 data platform for manufacturers. Designed a Spring Boot microservice that adds computational expressions dynamically for timely data processing, built its backend API to a Swagger spec, and set up Jenkins CI on GitHub.

  4. Oct 2016 — Apr 2017

    Beacon Innovation Group

    Android Developer Intern · Backend

    Patent search and drafting firm. Built the backend logic and REST API for querying patents in an Android patent-analysis app.

Skills

The AI and full-stack toolkit I ship with. Click any skill to see where I’ve used it.

Agents & orchestration

AI engineering

LLMs & prompting

AI engineering

Retrieval & RAG

AI engineering

Evals & observability

AI engineering

Frontend & mobile

Full-stack

Backend & APIs

Full-stack

Data & messaging

Full-stack

Cloud, DevOps & practice

Full-stack

Hiring for an AI engineering role?

I’m a senior full-stack engineer who ships LLM features to production, and I’m open to new roles. Email me, or ask Miso anything about my work first.

chiranjeevwork@gmail.com