Open to founding roles

Hi, I'm Brohammad

Founding Engineer

I ship production systems solo — from architecture to HTTPS.

Latest: VoxForge — open-source voice AI, live in production. I build the whole stack myself: pipeline, deploy, docs, and tests.

Systems in prod

Banking + VoxForge

Voice tests

354+

81% coverage

Transports unified

WS · REST · WebRTC

Deploy time

<15m

Docker + HTTPS

Open to founding rolesShips solo end-to-endBengaluruVoxForge live in prod

VoxForge Demo Resume

GitHub· 600+ commits LinkedIn

voxforge—zsh

prod

deploylogsmetrics

~ $git clone github.com/Brohammad/VoxFauge

~ $cd VoxFauge && ./deploy.sh init

✓ HTTPS ready · api, worker, postgres, redis, nginx

~ $curl -s https://voxforge.brohammad.tech/api/v1/health | jq

{"status":"healthy","transports":3,"coverage":"81%"}

~ $pytest -q --cov

✓ 354 passed · VoicePipelineService · all transports

~ $open https://voxforge.brohammad.tech/demo

✓ voice pipeline · STT → LangGraph → TTS → eval · ~100ms

~ $

voxforge-api · ubuntu-vps · 14d uptime~100ms mock

Scroll

About

Engineering profile

Who I am, why founding engineer, and what I'm building toward.

Who I am

Backend and AI engineer. By day, production systems at PwC. On my own time, I ship full platforms solo — most recently VoxForge, a voice AI stack live in production.

What I build

Production systems you can demo, deploy, and inspect — not slide decks. See Projects for the full stack; VoxForge is the latest and lives at voxforge.brohammad.tech.

What excites me

Zero-to-one ownership, voice AI infrastructure, and software that holds up when someone actually uses it.

Why founding engineer

I do my best work with full scope and full accountability. I want to be early — building the product, the infra, and the ops story together — not maintaining someone else's roadmap.

Why founding engineer

→I want to own problems end-to-end — architecture, implementation, deploy, and ops — not hand off to a platform team.
→Enterprise work taught me how production AI survives real load; I want that discipline at startup speed.
→I'm looking for a founding engineer or early co-founder role where shipping beats process.

Engineering Principles

First Principles

Understand systems instead of memorizing frameworks.

Production First

Build software that survives real users.

Continuous Learning

Always studying distributed systems, AI, and infrastructure.

Quality over Quantity

Fewer projects. Higher quality.

Observability

If you can't measure it, you can't improve it.

Reliability

Correctness before cleverness.

Beyond Code

Curious by default
Documentation enthusiast
Enjoy simplifying complex systems
Prefer elegant solutions over clever hacks
Build for maintainability

Open to

Open to co-founder conversations and early founding engineer roles — especially voice/AI infra. VoxForge is built in public so founders and engineers can evaluate how I work before we talk. Just shipped; no traction metrics yet.

Open toFounding engineerSide projectVoxForgeDay jobPwCFocusBackend + AILocationBengaluru

Current Focus

Founding Engineer RolesVoice AI InfrastructureZero-to-One ShippingProduction AIOpen SourceDistributed Systems

Currently Exploring

LiveKit WebRTCVoice Agent EvaluationCo-founder fitOpenTelemetryInference OptimizationSystem Design

Engineering Interests

Backend Engineering

Artificial Intelligence

Voice AI

Distributed Systems

Platform Engineering

Developer Experience

Open Source

Early-Stage Startups

Production Metrics

Production Systems

Enterprise + open source

Automated Tests

VoxForge test suite

Open Source Commits

GitHub contributions

~0K+

Daily API Requests

RAG + agent pipelines

Years Building

Professional engineering

Experience

Production engineering

Systems shipped to enterprise clients — architecture, scale, and impact documented.

Production

2025 — Present

Founder / Principal Engineer

VoxForgeBengaluru, India

Role

Solo founder

Scope

End-to-end

Status

Live in prod

Release

v1.0.0-rc.1

License

Open source

Site

voxforge.brohammad.tech

Production topology

Design

Build

Deploy

Docs

Release

SoloOpen sourceProductionHTTPS

What I Built

Founded VoxForge and shipped the full platform solo — from system design through production HTTPS
Automated deploy pipeline: fresh Ubuntu VPS to HTTPS in one command with Docker, NGINX, and Certbot
Published v1.0.0-rc.1 open source with ADRs, runbooks, and incident response documentation
Set CI quality gates: 354+ tests, Playwright UI coverage, dependency audit, and secret scanning

Architecture Decisions

→Scoped as modular monolith to ship pilots faster without a team
→Prioritized operator trust — eval, replay, handoff — over demo-only features
→Open-sourced to lower adoption barrier for pilot customers and contributors

Engineering Challenges

End-to-end ownership without a team
Shipping Vapi-class scope as a solo engineer
Making the project evaluable without a sales deck

Production Impact

Solo founder — architecture through production deploy
Open-source release with ADRs and runbooks
Live at voxforge.brohammad.tech
Built for recruiters, pilots, and eng evaluators
v1.0.0-rc.1 shipped on GitHub
No vendor lock-in or per-minute SaaS fees

Engineering Highlights

Scope

Architecture, backend, deploy, docs — solo

Ownership

Founder & principal engineer, end-to-end

Release

v1.0.0-rc.1 open source on GitHub

Audience

Pilot customers, recruiters, eng evaluators

Documentation

ADRs, runbooks, incident response docs

Quality bar

CI gates, coverage, secret scanning

Technology Stack

Core

PythonFastAPILangGraph

Voice

LiveKitWebSocketsSTT/TTS

Data

PostgreSQLpgvectorRedis

Infra

DockerNGINXOpenTelemetry

Current

Production

2025 — Present

Associate Consultant

PwCBengaluru, India

Production Systems

Enterprise Clients

Fortune 500

AI Workflows

Multi-Agent

Cloud

Azure

Deployment

Production

Role

Backend + AI

Production topology

Client

API

Agents

RAG

Store

Azure AKSPostgreSQLQdrantRedis

What I Built

Shipped 3 production AI systems to enterprise banking clients
Built multi-agent insurance renewal platform reducing manual processing
Deployed RAG assistant handling thousands of queries daily in production

Architecture Decisions

→LangGraph state machines over ad-hoc agent chains for reliability
→Hybrid retrieval (dense + BM25) over pure vector search for finance docs
→Eval pipelines as a deployment gate, not a post-launch afterthought

Engineering Challenges

Reliable agent handoffs without state loss
Hallucination guardrails on regulatory content
Legacy banking API integration with audit trails

Production Impact

3 production AI systems delivered
Enterprise banking deployment
Thousands of production RAG queries
Multi-agent orchestration
Production backend APIs
Cloud-native deployment on Azure

Engineering Highlights

Architecture

Multi-agent LangGraph + hybrid RAG on FastAPI

Backend

REST APIs, SSE streaming, Redis session state

LangGraph flows, eval pipelines, structured outputs

Cloud

Azure AKS, Azure OpenAI, managed services

Infrastructure

Docker, PostgreSQL, Qdrant, Redis

Deployment

Production on Azure with CI/CD gates

Technology Stack

Languages

PythonTypeScript

Backend

FastAPINode.jsRESTSSE

LangGraphLangChainLlamaIndexOpenAI

Cloud

AzureAKSAzure OpenAI

Data

PostgreSQLQdrantRedis

Infrastructure

DockerGitHub Actions

Education

B.Tech · Electrical & Electronics Engineering

National Institute of Technology Calicut

2021 – 2025

Projects

Production systems

Deployed software with documented architecture, scale, and engineering trade-offs.

WebSocket

REST

LiveKit

VoicePipelineService

STTmulti-provider

→

agentLangGraph

→

TTSmulti-provider

→

evalper-turn

✓ RAG · replay · handoff~100ms mock

ProductionOpen SourceLive DemoVoice AI

VoxForge

Voice AI infrastructure you can deploy, operate, and trust.

Built for teams evaluating self-hosted voice AI — try the live demo or read the repo without a sales call. Just shipped; growing from here.

Release

v1.0.0-rc.1

Tests

354+

Coverage

81%

Deploy

<15 min

Production Highlights

Live at voxforge.brohammad.tech
354+ automated tests · 81% coverage
v1.0.0-rc.1 · 3 unified transports

Engineering Challenge

Modular monolith over microservices — faster pilots, simpler ops, module boundaries in code

Stack

Backend

PythonFastAPIUvicorn

AI / Voice

LangGraphRAGLiveKit

Data

PostgreSQLpgvectorRedis

Infra

DockerNGINXOpenTelemetry

Visit VoxForgeGitHub Demo

Room photo

→

Shade preview

Preview<60s

Shades63

Tests200+

ProductionLive DemoSolo Build

PaintCRM

From room photo to painted preview, qualified lead, and receivables — offline-first at paintcrm.brohammad.tech.

Built for paint dealers on the shop floor — preview 63 real catalog shades on a customer's wall in under 60 seconds, capture leads, and run quotes even without reliable internet.

Preview time

<60 sec

Catalog

63 shades

Tests

200+

API

40+ endpoints

Production Highlights

Live at paintcrm.brohammad.tech
200+ automated tests
40+ API endpoints

Engineering Challenge

Heuristic wall segmentation with optional ML fusion in the browser

Stack

Frontend

Vanilla JSViteCanvas 2D

Backend

Node.jsExpress

Data

PostgreSQL

Infra

DockerRender

GitHub Demo

DSLYAML rules

→

backtestbar-by-bar

→

optimizeCelery

→

tradeAlpaca

Sharpe · drawdown · win rateOpenAI insights

Open SourceFull-StackFintech

AlphaEdge

Design strategies. Backtest history. Trade with confidence — YAML DSL to paper execution with AI-powered insights.

Built for indie quant researchers who want strategy authoring, backtesting, optimization, and paper trading in one workspace — without juggling notebooks, spreadsheets, and broker APIs.

Stack

FastAPI + React

Workers

Celery

Tests

80+

Broker

Alpaca

Production Highlights

80+ unit and integration tests
Multi-symbol backtests with slippage and commissions
OAuth + RBAC-secured REST APIs

Engineering Challenge

Bar-by-bar simulation with equity curve and trade log accuracy

Stack

Backend

PythonFastAPICelery

Frontend

ReactTypeScript

Data

PostgreSQLRedis

Integrations

AlpacaStripeOpenAI

GitHub

Policyagent

Riskagent

Outreachagent

LangGraphorchestratorstate machine

PostgreSQL

Redis

LLM

ProductionEnterpriseMulti-agent

RenewAI

Multi-agent insurance renewal platform orchestrating policy analysis, risk scoring, and client outreach.

Deployment

Azure

Agents

3 specialized

Availability

99.9% target

Database

PostgreSQL + Redis

Production Highlights

Fortune 500 banking client
Multi-agent architecture
Production on Azure

Engineering Challenge

Reliable agent handoffs without state loss

Stack

Core

PythonFastAPI

LangGraphLangChain

Data

PostgreSQLRedis

Infra

DockerAzure

GitHub

querybanking policy Q

→

hybridQdrant + BM25

→

rerank12 chunks

→

streamSSE + cite

✓ cited · RBAC enforced840ms p99

ProductionEnterpriseRAG

FinCore Banking Assistant

Enterprise banking assistant with hybrid retrieval, source citation, and RBAC over regulatory documents.

Volume

~5K req/day

Latency

p99 840ms

Cloud

Azure AKS

Retrieval

Hybrid dense + BM25

Production Highlights

Enterprise production deployment
Thousands of queries daily
Hybrid retrieval pipeline

Engineering Challenge

Hallucination guardrails on regulatory content

Stack

Core

PythonFastAPI

LangChainOpenAI

Data

QdrantPostgreSQL

Infra

AzureSSE

GitHub

follower

leader

follower

log replication

Raft · linearizable · gRPC

DistributedConsensusIn ProgressBuilding

Distributed Key-Value Store

Fault-tolerant distributed store with Raft consensus, linearizable reads, and automatic leader failover.

Systems learning project — Raft consensus from scratch to understand linearizable reads/writes, leader election, and fault tolerance firsthand.

Nodes

3-node cluster

Protocol

Raft

Guarantee

Linearizable

Transport

gRPC

Production Highlights

Raft consensus
Linearizable guarantees
3-node cluster

Engineering Challenge

Correct Raft term management at partition boundaries

Stack

Core

Distributed

RaftgRPC

Data

Log replication

Infra

DockerProtobuf

View all projects on GitHub

Capabilities

Engineering capability matrix

What kinds of systems I build — organized by domain, not syntax.

Production Experience

Backend Systems

Designing scalable APIs, authentication, caching, asynchronous workflows, and production services.

FastAPINode.jsRESTgRPCWebSocketsSSEMicroservicesKafka

Production Experience

Voice AI

End-to-end voice agent platforms — transport unification, STT/LLM/TTS pipelines, WebRTC, and operator tooling.

LiveKitWebSocketsLangGraphSTT/LLM/TTSPer-turn EvalSession ReplayHuman Handoff

Production Experience

AI Engineering

Building RAG pipelines, multi-agent orchestration, LLM evaluation, and production inference workflows.

RAGLangGraphLangChainpgvectorMCPHybrid RetrievalLLM Eval

Advanced

Infrastructure

Containerization, CI/CD, observability, and deployment pipelines for production services.

DockerNGINXGitHub ActionsLinuxOpenTelemetryPrometheusLet's Encrypt

Advanced

Data Systems

Relational and vector stores, caching layers, and data modeling for high-throughput applications.

PostgreSQLpgvectorRedisQdrantMongoDBVector Search

Advanced

Cloud & Deployment

Cloud-native architecture, container orchestration, and secure infrastructure on Azure and AWS.

AzureAWSKubernetesDockerNetworkingSecrets

Working Knowledge

Developer Experience

Tooling, testing, documentation, and automation that help teams ship reliably.

GitPytestReactNext.jsAutomationAPI Docs

Working Knowledge

Testing & Reliability

Eval pipelines, integration testing, and reliability patterns for production AI systems.

LLM EvaluationpytestPlaywrightCoverage GatesFailure-mode Tests

Currently Exploring

Separate from production capabilities — active areas of study.

GoRustDistributed SystemsKubernetesCompilersOperating SystemsNetworkingSystem Design

Tech Stack

Core production stack

The technologies behind shipped systems.

PythonFastAPILangGraphLiveKitPostgreSQLpgvectorRedisDockerTypeScriptNext.jsAzureOpenTelemetry

Now

Engineering journal

Updated July 2026. Active work, research, and what I'm thinking about.

Current Obsession

Unifying voice transport layers without duplicating business logic — one pipeline for WebSocket, REST, and LiveKit.

Favorite Problem

Closing the gap between voice demos and production operability: evaluation, replay, and human handoff.

Currently Building

VoxForge v1.0 — voice AI platform live at voxforge.brohammad.tech
Distributed Key-Value Store — Raft consensus in Go
Per-turn evaluation suites as CI deployment gates

Currently Reading

Designing Data-Intensive Applications — Kleppmann
VoxForge ADRs and incident response runbooks

Currently Exploring

LiveKit WebRTC worker integration patterns
OpenTelemetry instrumentation for voice pipelines
MCP tool routing in production agents
Inference optimization for real-time STT/TTS

Current Research

Mock provider strategies for zero-API-key CI and demos
pgvector retrieval tuning for voice agent grounding
Failure-mode testing for WebSocket lifecycle and handoff flows

Next Projects

Event-driven notification service with Kafka
API gateway with rate limiting and observability

Engineering Goals

Ship fewer systems, each production-grade
Grow VoxForge pilot adoption and contributor base
Publish technical writing on voice AI infrastructure

Ideas

Signed replay links as regression fixtures in CI
Policy presets as operator-configurable agent guardrails
Eval suites tied to CI, not slide decks

Technical Notes

Single VoicePipelineService eliminated transport-specific test duplication
Mock providers complete voice pipeline in ~100ms without API keys
Modular monolith shipped faster than planned microservices split

Uses

Engineering setup

Tools I reach for daily when building production systems.

Hardware

MacBook Air

Primary development machine

Editor

Cursor

AI-first code editor

VS Code

Secondary editor

Terminal

Warp

Modern AI-powered terminal

zsh + Oh My Zsh

Shell config

AI Tools

Cursor

AI-first coding

Claude

Technical reasoning and writing

ChatGPT

General purpose

Perplexity

Research and documentation

Dev Tools

Docker

Containerization

Postman

API testing

TablePlus

Database GUI

GitHub

Version control

Open Source

GitHub

Systems work, experiments, and production patterns — building in public.

@Brohammad

Backend, AI, and distributed systems

600+ commits

bank-fincorp

Enterprise RAG assistant with hybrid retrieval, source citation, and SSE streaming for banking workflows.

Python

InsuranceAI

Multi-agent insurance renewal platform orchestrating policy analysis, risk scoring, and client outreach.

Python

distributed-kv-store

Raft consensus implementation with linearizable reads and gRPC inter-node communication.

auth-service

JWT authentication microservice with refresh rotation and RBAC.

TypeScript

url-shortener-service

High-throughput URL shortener with Redis caching and rate limiting.

TypeScript

View GitHub Profile

Writing

Technical notes

Deep dives on production backend and AI engineering. In progress.

Articles in progress

Writing on systems that survive production — not tutorials.

Planned topics

Building reliable multi-agent systems with LangGraph
RAG in production: chunking strategies that actually work
Implementing Raft consensus from scratch in Go
Redis caching patterns for high-throughput APIs
Prompt engineering for structured outputs

+3 more topics queued

Contact

Get in touch

Open to founding engineer roles and early co-founder conversations. Founders, recruiters, and engineers welcome.

If you're a founder building voice agents, AI infra, or developer tools — or hiring a founding engineer who ships solo — I'd like to hear from you.

aabidmohamed2003@gmail.com

Location

Available Globally · Open to Relocation

Connect

GitHub LinkedIn Email Resume