Available for freelance & contract work

I build AI that does the
work people don't want to do.

I'm Anshul — an engineering lead with 7+ years building autonomous, agentic AI systems that automate slow, document-heavy work. I help companies ship AI that survives production and cuts real cost — not just demos.

80% fewer client man-hours Scaled systems to 9M+ users
Understand Prototype Harden Hand off
0
Years building
production software
0
Client man-hours
cut with AI
0
Users scaled
across products
0
Microservices run
in production
0
Peak daily
active users
AG

Anshul Goel

Agentic AI Engineer · Engineering Lead
  • Currently leading AI at ReBillion.ai
  • Built systems serving 9M+ users
  • Works with U.S. clients, remote worldwide
  • Bangalore, India
who i am

Every job has a boring 80% — the repetitive busywork nobody signed up for. I build AI agents that do that part.

For the last few years I've gone deep on autonomous, agentic AI. At ReBillion.ai I led the engineering team building an AI transaction coordinator for U.S. real estate — systems that read messy legal paperwork and act on it, cutting client man-hours by ~80% so they scale without hiring more people.

But agentic AI sits on a decade of building backends that don't fall over. Before the LLM wave I scaled platforms to 9M+ users and a peak 250K+ daily actives, ran 100+ microservices across AWS, GCP and Azure, and took products to real revenue.

What that means for you: an engineer who connects AI to outcomes — and ships things that actually work in production, not just in a demo.

services

What I can build for you

From a stuck AI prototype to a workflow buried in PDFs — here's where I help, and who each is for.

Agentic AI Systems

Autonomous multi-agent workflows (LangGraph, Vertex AI) that plan, decide and act end-to-end — with a human only where it matters.

For teams who want AI that does the task, not just chats

Document Intelligence & OCR

Turn messy PDFs, scans and forms into clean, validated, structured data — with AWS Textract, Google Document AI, Mistral & LlamaParse.

For ops-heavy businesses drowning in paperwork

GenAI / LLM Product Engineering

Take AI features from a flashy prototype to dependable production — evals, guardrails, reliability and the unglamorous engineering that makes it trustworthy.

For founders whose AI needs to work in front of real users

Backend & Distributed Systems

APIs and infrastructure that stay fast and reliable at millions of users — Golang, Node.js, multi-cloud, microservices, queues and observability.

For products that need to scale without breaking

Fractional AI Engineering Lead

Senior architecture, technical direction and team setup for companies adding AI to their product — without the cost of a full-time hire. I help you choose the right approach, de-risk it, and get your team shipping.

For startups that need senior AI leadership, part-time
how i work

From idea to something that runs itself

A simple, low-risk path — so you see value before you commit to the whole build.

Understand

We dig into the real problem and the workflow behind it — not just the feature request.

Prototype

A working proof-of-concept, fast — so we validate the approach before investing in it.

Harden

Engineering for production: reliability, evals, edge cases and scale. The part most demos skip.

Hand off

Clean docs and knowledge transfer so your team can run and extend it confidently.

selected work

Proof, not promises

Two systems I led — what the problem was, how I approached it, and what it delivered.

ReBillion.ai · Agentic AI

Autonomous AI for U.S. Real Estate

Engineering Lead · 2023–present
Challenge
Real-estate transactions are buried in manual coordination — chasing signatures, checking dates, sorting legal PDFs nobody wants to read.
Approach
Led the team building an agentic AI control plane (LangGraph, Vertex AI) on top of an OCR/IDP stack (Textract, Document AI, Mistral, LlamaParse) that reads the paperwork and acts on it — validating files, flagging errors, moving deals forward.
Result
~80% fewer client man-hours. Clients close more deals without scaling headcount.
80%
manual coordination work removed
Multi-agent
LangGraph · Vertex AI
4 OCR engines
Textract · Doc AI · Mistral · LlamaParse
Leher · Scale & Systems

Scaling a Social Platform to Millions

Intern → Tech Lead · 2018–2023
Challenge
Build and scale audio-video social products for a fast-growing, massive audience — reliably and cheaply.
Approach
Architected 100+ microservices across AWS, GCP & Azure, a push system doing 10M+ notifications/day, and led a team of 5 — leveraging startup credits to keep cloud costs near zero.
Result
9M+ users, peak 250K+ DAU, a $100K/month service and a product at 10cr ARR on $0 cloud spend.
9M+
users served across products
250K+
peak daily actives
10cr
ARR · $0 cloud cost
the toolkit

From idea to production

The technologies I reach for across the agentic-AI stack.

Agentic AI

LangGraphVertex AIAI AgentsRAGMulti-agent

Document Intelligence / OCR

AWS TextractGoogle Document AIMistralLlamaParseIDP

Backend & Distributed

GolangNode.jsMicroservicesQueuesObservability

Cloud & DevOps

AWSGCPAzureDockerCI/CD
why me

Senior, end-to-end, no hand-holding

Outcomes, not demos

I optimise for what moves your business — cost, time, revenue — and ship things that survive real users.

Battle-tested at scale

Millions of users, 100+ microservices, multi-cloud. I've seen what breaks — and how to keep it from breaking.

Senior ownership

I lead engineering teams day-to-day. You get someone who owns the problem, plans it, and drives it to done.

Clear communication

Plain-English updates, honest timelines, and no surprises. You always know where the project stands.

Let's build something
that runs itself.

Got a workflow that should be automated, or an AI idea stuck at the demo stage? Tell me about it — the first call is on me.