Open to AI engineering roles

Tiju Raju Lukose · AI Engineer

I build AI agents and RAG systems that hold up in production.

2+ years at Hexaware Technologies shipping FastAPI services, retrieval systems and LLM agents for financial services clients. I work on what makes AI reliable: evaluation, guardrails, observability and cost.

Stack I ship with

0k
API requests a day on services I built
0%
uptime on AWS ECS through 3x traffic spikes
0%
lower token spend per query from retrieval
0
AI systems built end to end

Core technologies

Agentic AI

  • LangGraph
  • LangChain
  • MCP
  • Multi-agent systems
  • Tool calling
  • Agent memory
  • Human-in-the-loop
  • Guardrails
  • CrewAI

LLMs and retrieval

  • OpenAI
  • Azure OpenAI
  • Anthropic Claude
  • RAG
  • Hybrid search
  • Reranking
  • GraphRAG
  • pgvector
  • Pinecone
  • FAISS
  • Structured outputs

Evals and reliability

  • Golden datasets
  • LLM graders
  • Regression evals
  • LangSmith
  • OpenTelemetry
  • Red teaming
  • Cost and latency tuning

Model engineering

  • Hugging Face
  • LoRA / QLoRA
  • PEFT
  • vLLM
  • Open-source models
  • Vision-language models
  • Voice AI

Backend and frontend

  • Python
  • TypeScript
  • FastAPI
  • Pydantic
  • Node.js
  • PostgreSQL
  • Redis
  • MongoDB
  • React
  • Next.js

Cloud and DevOps

  • AWS (ECS, Fargate, Lambda, RDS, SQS)
  • Azure
  • Docker
  • Kubernetes
  • Terraform
  • GitHub Actions

Selected work

End-to-end AI systems, each with an evaluation suite, tests and CI.

Flagship Health insurance

Cashless Claims Copilot

Agentic copilot for a hospital insurance desk. Turns admission and discharge records into insurer-ready cashless claims, predicts deductions before submission and answers insurer queries with cited evidence. Every action on an insurer portal goes through human approval.

  • LangGraph
  • MCP
  • FastAPI
  • pgvector
  • Next.js
  • AWS
  • Terraform

Flagship Pharma compliance

Data Integrity Copilot for Pharma QC

Audit-trail analytics and investigation agent for quality-control labs. Flags data integrity red flags across chromatography and LIMS systems, drafts deviation reports with cited evidence and prepares sites for FDA inspections. Runs on premises with open models.

  • LangGraph
  • Open models
  • Fine-tuning
  • FastAPI
  • React
  • Kubernetes

Flagship Privacy

Indic PII Shield and DPDP Rights Desk

API, SDK and MCP server that finds and redacts Indian personal data in 10+ languages, code-mixed text and ID-card images before it reaches an LLM. An agent handles DPDP access, correction and erasure requests across company systems.

  • MCP
  • Transformers
  • FastAPI
  • PostgreSQL
  • Kubernetes
  • Azure

Text-to-SQL Analyst Agent

Ask a Postgres database questions in plain English. The agent finds the right schema, writes and safety-checks SQL, runs it read-only, fixes its own errors and returns a table, chart and explanation.

Benchmarked on BIRD Mini-Dev with ablations

  • LangGraph
  • pgvector
  • FastAPI
  • Next.js

LLM Gateway

OpenAI-compatible proxy in front of many model providers with per-key budgets, rate limits, cost-aware routing, retries, circuit breakers, fallbacks, and exact plus semantic caching.

Load-tested with k6, chaos-tested for provider outages

  • FastAPI
  • Redis
  • Kubernetes
  • OpenTelemetry

GST Invoice Extractor

Indian GST invoices from scans, photos and PDFs to validated JSON with GSTIN checksum and tax arithmetic checks. A LoRA fine-tuned open vision-language model, compared with frontier APIs on accuracy, cost and speed.

Field-level F1, cost per 1k docs, p95 latency

  • PyTorch
  • PEFT
  • vLLM
  • Next.js

Issue-to-PR Coding Agent

A GitHub App. Label an issue and a sandboxed agent reproduces the bug with a failing test, patches the code, runs the suite and opens a draft PR with evidence.

Evaluated on SWE-bench Verified Mini

  • Python
  • Docker
  • GitHub Apps
  • Redis queues

Agent Memory Service

Long-term memory any agent can plug into over REST or MCP. Extracts typed memories, tracks when facts change, retrieves with hybrid, recency and entity search, and fully erases a user's data on request.

Evaluated on LongMemEval

  • MCP
  • PostgreSQL
  • pgvector
  • FastAPI

Realtime Voice Receptionist

Phone and web voice agent that books, moves and cancels clinic appointments in Hindi, English, Marathi and code-mixed speech, with barge-in, a per-turn latency budget and safety escalation.

160 simulated-caller runs across 4 language modes

  • Pipecat
  • Sarvam
  • FastAPI
  • Next.js

Also built

Where I've worked

Hexaware Technologies

Navi Mumbai, India
May 2024 to present

  1. Software Engineer-2Jul 2026 to present
  2. Software Engineer-1May 2025 to Jul 2026
  3. Associate Software EngineerSep 2024 to May 2025
  4. Premier Graduate Engineer TraineeMay 2024 to Sep 2024
  • Put OpenAI, Azure OpenAI and Hugging Face models behind one service with retries, fallbacks and cost-aware routing. Retrieval over 25,000 documents cut token spend per query by 40%.
  • Built FastAPI microservices for RAG document search: 50 endpoints across 5 services serving 70,000 requests a day.
  • Cut API latency by 13% and database read load by 28% with async I/O, query tuning and a Redis cache.
  • Ran containerised services on AWS ECS and Fargate at 99.9% uptime through 3x traffic spikes.
  • Took releases from 3 days to 15 minutes with GitHub Actions and 95% test coverage.
  • Led code reviews and mentored 2 junior developers to independent feature ownership.

Recognition

  • Winner, Shark Tank for GenAI in Financial Services hackathon, Hexaware (2024)
  • Winner, EcoFuelers hackathon, Hexaware (2025)
  • 500+ problems solved on LeetCode

Education and certifications

  • B.E. Computer Science and Engineering, Indira College of Engineering and Management, Pune
  • Microsoft Certified: Azure Fundamentals
  • Generative AI and LLM Application Development, AI Accelera

Let's build something reliable.

Hiring for AI engineering roles? I'm open to Pune, hybrid and remote positions.