rs.
Resume

RAHUL SAINI / AI / ML ENGINEER

From data.
To intelligence.
Into production.

I build AI systems that work beyond the demo.
Data science rigor meets production engineering,
from RAG pipelines to conversational agents.

Explore my work View resume
Based in Irvine, CA · Building on AWS
SYSTEMS THINKING / 001
APPLIED INTELLIGENCE
Data
Reasoning
Production
DESIGNED TO CONNECT. BUILT TO WORK.
Senior AI/ML Engineer TCS @ Hyundai Capital America MS, Data Science University of Alabama at Birmingham AWS Certified AI Practitioner

01 / SELECTED WORK

Ideas, engineered
for the real world.

Enterprise systems and personal builds.
Thoughtful architecture. Useful outcomes.

CONVERSATIONAL AI · ENTERPRISE01

NL2SQL Conversational Agent

A question in. An answer out.

Turning plain-English questions into database queries with LangChain and AWS Bedrock, bringing business teams closer to their data.

LangChainAWS BedrockNL2SQL
In development · TCS
Project notes

An agent designed to reduce insight delivery from weeks to minutes and eliminate reliance on dedicated BI resources. Enterprise source code is not publicly available.

FULL-STACK AI · PERSONAL PROJECT02

FinanceOS

A clearer picture of your money.

A personal finance platform with Maya, an embedded AI financial advisor with structured responses and behavioural guardrails.

Next.jsTypeScriptSupabaseClaude API
Shipped in 10 days
Project notes

Delivered 20+ pages, 15+ API routes, bank integrations, and cron automation. A TypeScript prompt template suite defines Maya’s persona, output formats, and runtime variables. Built with an AI-assisted development workflow and reviewed outputs.

AWS-NATIVE SYSTEMS · ENTERPRISE03

Fleet as a Service

The right vehicle. The right dealer.

An autonomous onboarding platform matching vehicles to dealers by capacity and proximity, using Haversine and real driving distance.

AWS LambdaStep FunctionsPython
100–500 VINs processed weekly
Project notes

The geospatial engine assigns grounded vehicles across 2,000+ dealers. Step Functions and Lambda orchestrate the onboarding workflow, cutting dealer assignment turnaround from days to hours. Enterprise source code is not publicly available.

AGENT AUTOMATION · PERSONAL PROJECT04

AI Research Newsletter Agent

Research, without the noise.

A weekly research pipeline that searches, filters, and turns AI developments into a useful Monday briefing.

n8nGPT-4.1TavilyJavaScript
Automated weekly delivery
Project notes

The agent produces structured JSON, while a dedicated JavaScript node handles HTML email rendering. A signal-quality heuristic filters opinion pieces and duplicates across seven categories. Each story includes a factual summary, a 30-second spoken briefing, and an engineering takeaway.

The learning never stops.Explore my Gen AI Cookbook: foundations, experiments, and applied projects.

02 / THE FOUNDATION

An engineer’s mindset.
A data scientist’s rigor.

My path to AI engineering started with data: building pipelines, asking better questions, and learning to trust the evidence.

Today, I’m a Senior AI/ML Engineer at TCS, working with Hyundai Capital America on production GenAI systems. My MS in Data Science from UAB grounds that work in statistics, evaluation, and careful reasoning.

I bring that foundation to every layer, from the data underneath an agent to the cloud infrastructure that keeps it running.

The journey so far

AI & LLM engineering

LangChain · LangGraph · RAG pipelines · NL2SQL · Prompt engineering · LLMOps

Cloud & infrastructure

AWS Bedrock · Lambda · Step Functions · Glue · S3 · Docker · CI/CD

Data & ML foundations

Python · SQL · Pandas · NumPy · Apache Spark · PostgreSQL · Tableau

Full-stack development

React · Next.js · TypeScript · Node.js · Supabase · REST APIs

03 / EXPERIENCE

Built on
experience.

From full-stack development and data platforms to applied AI.

Read the full resume

Jul 2025 – Present CURRENT

AI/ML Engineer

Tata Consultancy Services

  • Architected an end-to-end autonomous vehicle onboarding platform processing 100–500 VINs weekly using AWS Step Functions and Lambda, cutting dealer assignment turnaround from days to hours.
  • Built a geospatial dealer matching engine (Haversine + real driving-distance API) that optimally assigns grounded vehicles across 2,000+ dealers by capacity and proximity.
  • Developing an NL2SQL AI agent powered by LangChain and AWS Bedrock that converts plain-English questions into database queries, reducing insight delivery from weeks to minutes and eliminating dedicated BI resources.

Aug 2021 – Jun 2025

Business Intelligence Developer

Tata Consultancy Services

  • Designed serverless orchestration architecture with AWS Step Functions and Lambda, delivering $40K+ in annual cloud cost savings through optimised workflow execution.
  • Automated ETL pipelines via AWS Glue, Informatica, and Python, eliminating 90% of manual data handling and accelerating reporting cycles for the FP&A team.
  • Delivered a document intelligence PoC using AWS Textract and Lambda to extract transactional data from financial PDFs into structured databases, replacing a fully manual process.

Mar 2021 – Aug 2021

Full Stack Developer Intern

TA Digital

  • Built reusable React.js components with REST API integration and Spring Boot microservices handling authentication, CRUD operations, and PostgreSQL persistence with role-based access control.
  • Contributed to CI/CD pipeline setup using GitHub and Docker in an Agile team, reducing post-deployment bug rate through JUnit and Postman test suites.

THE FULL PICTURE

A little more detail.

Experience, projects, education, and the tools behind the work.

Download resume PDF

04 / LET’S CONNECT

Good systems start
with a conversation.

Have an AI problem worth solving? Let’s talk.