ms
Hire me
Available · Q3 2026
Portfolio · Selected Works

Manjot Singh

Currently
AI Solutions Architect
Optimum Data Analytics
Co-founder
CTO
TechnoTribes
Building
Founder
Buildify

I architect agentic AI systems, ship voice agents, and build the developer infrastructure that turns prototypes into products.

Seven years across Reliance, TATA Digital, and SaaS clients in the US and India. Now leading AI delivery at ODA while operating two product companies on the side.

See products → Start a project
10+
years shipping
3
active clients
20+
products in market
∞
side bets running
scroll
◆ Multi-agent systems
◆ Voice AI · agents
◆ RAG pipelines
◆ LLM orchestration
◆ Agentic automation
◆ React · Next.js · Remix
◆ Node · NestJS · Python
◆ Postgres · Mongo · pgvector
◆ AWS · GCP · Azure
◆ Docker · Kubernetes
◆ Vercel AI SDK
◆ Anthropic · OpenAI
◆ PWA at super-app scale
◆ Multi-agent systems
◆ Voice AI · agents
◆ RAG pipelines
◆ LLM orchestration
◆ Agentic automation
◆ React · Next.js · Remix
◆ Node · NestJS · Python
◆ Postgres · Mongo · pgvector
◆ AWS · GCP · Azure
◆ Docker · Kubernetes
◆ Vercel AI SDK
◆ Anthropic · OpenAI
◆ PWA at super-app scale
01 / Products

Six products.
One operator.

From multi-agent reasoning pipelines to voice agents in production — the work I've shipped across ODA, TechnoTribes, and my own ventures. Heavy on agentic AI, automation, and voice.

01
In production

Pre Sales Agent

Multi-agent presales intelligence

Optimum Data Analytics / AI Solutions Architect

A multi-agent system that ingests RFPs, qualifies opportunities, drafts solution architectures, and assembles presales decks in minutes. Replaces a 3-day analyst loop with an orchestrated reasoning pipeline running across specialised agents — research, qualification, scoping, and writeup.

▪ Multi-agent orchestration
▪ RFP ingestion + analysis
▪ Solution architecture drafting
▪ Auto-generated presales decks
PythonLangGraphAnthropicpgvectorFastAPI
02
Live · iq.technotribes.org

InterviewIQ

AI interview platform with voice agents

TechnoTribes / CTO · Founder

Drop-in AI interviewing infrastructure for technical hiring. Live voice agents conduct structured interviews, score candidates against rubrics, and generate transferable reports. Pairs with TechnoTribes recruitment to scale candidate evaluation without losing signal — Next.js + Postgres + Prisma + Claude.

▪ Live voice interview agents
▪ Async screening flows
▪ Rubric-based scoring
▪ ATS-ready reports
Next.jsPostgresPrismaClaude APIVoice AI
03
Client deployments

VoiceIQ + Petal

AI voice screening · customer support agents

TechnoTribes / CTO · Architect

Production voice agents for two distinct workflows. VoiceIQ handles outbound candidate screening at the top of recruitment funnels — the same flavor of voice tech we pitched to Zepto. Petal is an inbound customer support agent built for Ferns N Petals, handling order tracking, gifting flows, and floral curation queries with function calling.

▪ Outbound screening (VoiceIQ)
▪ Inbound support (Petal)
▪ Function-calling integrations
▪ Real-time transcription + handoff
Voice AIFunction callingRAGWebRTC
04
v2 in development

Buildify

AI-powered developer platform

Founder / Founder · buildify.sh

A platform layer for developers building AI-native products. Three primitives shipped so far: an Accessibility Engine that audits and remediates UIs, an AI Resume Builder with LaTeX-based PDF generation, and TestFish — a web testing engine that reasons about user flows instead of brittle selectors. Filed under the Arch Grants Startup Competition 2026.

▪ Accessibility engine
▪ AI resume builder (LaTeX)
▪ TestFish — agentic e2e testing
▪ Dev primitives + APIs
Next.jsPlaywrightLLM agentsTSLaTeX
05
Self-hosted

Prism

Social media scheduling tool

Indie / Solo build

Multi-channel post scheduling, content calendar UI, and AI-assisted caption rewriting using the humanizer framework — handles the four-week LinkedIn calendars I run for my own brands.

▪ Multi-channel scheduling
▪ Content calendar UI
▪ AI caption rewriting
▪ Humanizer framework built-in
Next.jsPostgresBull queueOpenAI
06
Technical plan complete

NSE/BSE Quant Platform

TradingView alternative for Indian markets

Indie · in design / Architect

A from-scratch platform targeting NSE/BSE — scoped to handle data licensing, intraday tick streams, custom indicator scripting, and backtesting. Quietly betting that Indian indie quants will need infrastructure that TradingView never prioritised. Phased technical roadmap and stack recommendations documented.

▪ NSE/BSE data licensing model
▪ Real-time tick streaming
▪ Pine-style indicator scripting
▪ Backtesting + paper trading
TBDLikely Next + RustTimescaleDBWebSockets
▸ Plus side projects in clinical data (CDISC/SDTM), Salesforce vertical AI, government CV systems, and Pine Script indicators.
02 / Capabilities

What I build
for clients.

Six engagement shapes. Each one mapped to real work I've shipped — not aspirational consulting fluff. Heavy on AI, voice, and agentic automation, with classical SaaS as the foundation.

01
01 / 06

Agentic AI systems

Multi-agent architectures that replace manual analyst work — research, reasoning, scoping, writeup. The Scoopy pattern, repeatable for any vertical: pharma, recruitment, presales, ops.

  • ▪ Agent orchestration (LangGraph)
  • ▪ Tool-use + function calling
  • ▪ Memory + state management
  • ▪ Eval harnesses
02
02 / 06

Voice agents in production

Inbound and outbound voice automation — candidate screening, customer support, qualification calls. The same engine I built behind VoiceIQ and Petal, hardened for real client volume with handoff and transcription baked in.

  • ▪ Outbound screening flows
  • ▪ Inbound support agents
  • ▪ Real-time transcription
  • ▪ Human handoff routing
03
03 / 06

AI integrations on existing stacks

Adding intelligence to products you've already built without a rewrite — RAG pipelines on top of CMS data, copilots inside dashboards, automation layers under existing workflows. Pragmatic, fast, measurable.

  • ▪ RAG pipelines
  • ▪ In-product copilots
  • ▪ Workflow automation
  • ▪ Embedding + vector search
04
04 / 06

SaaS platforms, end-to-end

Multi-tenant SaaS with secure billing, role-based access, and dashboards that hold up under load — the work I shipped for Simsi, CheQ, and OrbitalCustoms, repeatable for funded teams who need senior eyes on architecture.

  • ▪ Greenfield builds
  • ▪ Admin tooling
  • ▪ Billing flows
  • ▪ Enterprise auth
05
05 / 06

Enterprise frontend at scale

Frontend systems for products with millions of users. I built mobile-and-web inside one of India's largest super-app launches — every screen ships through review boards, analytics, microservice integrations. I know what production-grade actually means.

  • ▪ React / Next / Vue / Angular
  • ▪ PWA architecture
  • ▪ Microservice integration
  • ▪ Performance budgets
06
06 / 06

Technical advisory + fractional CTO

For founding teams making early stack bets — architecture reviews, hire planning, technical due diligence, AI roadmap design. Limited slots, paid by retainer or per-engagement.

  • ▪ Architecture reviews
  • ▪ AI roadmap design
  • ▪ Hiring + interviews
  • ▪ Due diligence
03 / Experience

Where I've
actually built.

Two enterprise giants. One contract studio. Two of my own ventures. Plus an AI architect role that ties them together. The track record under the title.

2025 — Now
Full-time

Optimum Data Analytics

AI Solutions Architect

Architecting AI delivery across multi-agent presales tools, clinical data platforms, and Salesforce vertical AI. Leading agentic automation initiatives end-to-end.

  • ▪ Architected presales agent — multi-agent presales intelligence system replacing 3-day analyst loops.
  • ▪ Designed OneData — clinical data platform aligned with CDISC/SDTM regulatory standards.
  • ▪ Built ORBIT — Salesforce vertical AI for regulatory consulting firms (bid qualification).
  • ▪ Contributed to Vision Central Belt — government CV system pitched to MeitY.
PythonLangGraphAnthropicSalesforceCDISC
2025 — Now
Founder

TechnoTribes

CTO · Co-founder

Operating a software studio + technical recruitment firm. Building AI products as flagship offers — voice agents, interview infrastructure, customer support automation.

  • ▪ Shipped InterviewIQ (live on iq.technotribes.org) — Next.js + Postgres + Claude.
  • ▪ Architected VoiceIQ outbound screening agents and Petal inbound customer support agents.
  • ▪ Lead engineering across all client deliverables and own the technical hiring loop.
Next.jsVoice AIPostgresClaude API
Jun 2025 — Jan 2026
Contract

Fayredge (APS Tech Pvt Ltd)

Technical Lead, Engineering

Leading engineering on a SaaS platform that handles surveys across multiple stakeholders and helps companies measure ROI.

  • ▪ Led end-to-end feature development across the stack — tracking millions of app surveys.
  • ▪ Reduced discrepancies in survey attribution by over 30%.
  • ▪ Improved billing success rates and cut transaction failures by 25%.
Full-stackScalable backendBilling flowsTypeScriptNestJSNext.jsReact
Oct 2023 — June 2025
Multi-client

Lauren Information

Senior Full-Stack Engineer

Three concurrent client engagements — US-based admin tooling, Indian fintech credit infrastructure, and a logistics SaaS.

  • ▪ Simsi (US) — Spearheaded a Remix + TypeScript admin tool with reusable UI components and secure backend integrations.
  • ▪ CheQ — Built React + NestJS dashboards for a credit bill payment platform, RBI-compliant with secure account management.
  • ▪ Lupin — Architected a Next.js SaaS; cut transaction failures by 25% via CI/CD and billing redesign.
RemixTypeScriptNestJSNext.jsReact
Oct 2019 — Oct 2023
TATA Neu

TATA Digital

Senior Full-Stack Engineer

Built consumer-facing surfaces inside one of India's largest super-app launches — TATA Neu.

  • ▪ Built mobile and web frontends on a PWA stack — Node.js, Python, React.
  • ▪ Integrated microservices across the super-app ecosystem; deployed on Linux via Azure DevOps.
  • ▪ Recognized twice in The Incredibles — TATA Digital R&R Program (Oct 2020, Oct 2021).
ReactNode.jsPythonPWAAzure DevOps
Aug 2018 — Oct 2019
Data Engineering

Reliance Industries

Full-Stack Engineer

Worked on automation tooling for one of the largest data engineering organizations in India.

  • ▪ Automated Spark scripts for HQL queries.
  • ▪ Built a UI for the DE framework — Angular frontend, Spring Boot microservices, MySQL backend.
  • ▪ Created and maintained end-to-end admin panel pages.
AngularSpring BootMySQLSparkHQL
▸ Recognition
The Incredibles
TATA Digital R&R · 2020 + 2021
▸ Recognition
Gold Medalist — Bhangra
Inter-University · 2015–2016
▸ Leadership
Head Boy of School
2013–2014
04 / Lab

The archive,
still warm.

Repos I've shipped, forked, prototyped, or kept in rotation — the longer tail behind the headline products. Coding agents, voice surfaces, dev infra, and a personal assistant I run on my own devices.

▸ 01 · Track

Blackbox AI · Coding agents suite

Engineering across the Blackbox AI product surface — from a 3M+ install IDE extension to the LLM router behind it.

BLACKBOXAI Agent (VSCode)

repo

Autonomous coding extension · 3M+ installs

Production VSCode extension that creates and edits files, runs commands, and drives a browser with permission. Shipped on the Blackboxapp publisher.

TypeScriptVSCode APIMCPClaude · GPT

Blackbox CLI

repo

Terminal coding agent · npm workspace

Node 20+ CLI agent with sandboxed execution (Docker/Podman), VSCode companion, and integration tests against terminal-bench. Built around a parser tuned for Blackbox models.

Node.jsTypeScriptesbuildVitestDocker

Blackbox AI Code Chat

repo

Web AI chat · Vercel AI SDK

Next.js App Router chatbot with streaming UI, multi-provider model switching, KV-backed history, and rate limiting. Integrates Sandpack, Liveblocks, LiveKit, and ElevenLabs for richer flows.

Next.jsVercel AI SDKAnthropicUpstash KVNextAuth

AI Builder (v0 clone)

repo

Multi-tenant app builder

A v0-style AI Builder built on the v0 SDK + AI Elements with Drizzle, Neon Postgres, Stripe, and Statsig. Generates and previews full Next.js apps from prompts.

Next.jsv0 SDKDrizzleNeonStripe

Blackbox AI Router

repo

LLM proxy · model routing

Python proxy server fronting multiple model providers via LiteLLM + FastAPI. Container-first deploy, Pydantic-validated config, dotenv-driven keys per environment.

PythonLiteLLMFastAPIPydanticDocker

OI Proxy Server

repo

OpenInterpreter routing

Companion proxy + prompt/model config used by the Blackbox OpenInterpreter integration. LiteLLM proxy with provider keys and routed planner/executor models.

PythonLiteLLMDocker
▸ 02 · Track

Vercel · sandboxed agent platforms

Two coding-agent harnesses built around Vercel Sandbox — one a multi-agent template, one a vibe-coding playground.

Coding Agent Template

repo

Multi-agent runner · Vercel Sandbox

Template that runs Claude Code, Codex CLI, and Blackbox CLI inside isolated Vercel Sandboxes. NextAuth + GitHub for repo access, Twilio SMS on completion, AI-generated branch names.

Next.jsVercel SandboxAI GatewayDrizzleLiveKit

Vibe Coding Platform

repo

Prompt-to-app agent

Next.js coding agent on Vercel Sandbox with Monaco, Octokit-backed GitHub flows, Stripe billing, Playwright in-loop, and Arctic-based OAuth. Generates and runs apps directly from chat.

Next.jsVercel SandboxAI SDKMonacoStripe
▸ 03 · Track

Personal · indie infra

My own runtime — what I run on my devices, and the context layer feeding everything else.

Clawdbot

repo

Personal AI assistant · 9 channels

Single-user assistant on WhatsApp, Telegram, Slack, Discord, Google Chat, Signal, iMessage, Teams, and WebChat — plus BlueBubbles, Matrix, and Zalo. Speaks and listens on macOS/iOS/Android, with a live Canvas you control. Pluggable via a Plugin SDK.

TypeScriptBaileysPlugin SDKOAuth (Anthropic · OpenAI)

Context-Man (MCP)

repo

MCP server · project context

MCP server that gives AI assistants structured access to PLAN/TODO/SCOPE/TECHSTACK, a skills library, code snippets, and prompt templates. Ships CLI + REST + MCP interfaces, MongoDB-backed, full-text search, version history.

TypeScriptMCP SDKMongoDBExpressCommander
▸ 04 · Track

Client · vertical builds

Production work for client engagements — surveys, voice support, regulated forms.

Orbit Validator

repo

Validator UI

Next 16 + React 19 validator surface paired with the Orbital install pipeline — Tailwind v4, App Router, focused on schema and rule checks before submission.

Next.js 16React 19Tailwind v4

FNP Support

repo

Ferns N Petals · Petal POC

React + Vite POC surface that fed into the Petal inbound support agent — order tracking, gifting flows, and floral-curation queries scoped before the production voice build.

React 19ViteJS
▸ 05 · Track

OSS · explorations

Open-source contributions and POCs kept on the bench.

Config Manager

repo

JSON Schema · dynamic forms

React POC for a config UI generator built on @rjsf with MUI bindings. Used to prototype the form-driven configuration patterns later folded into client work.

React 19react-jsonschema-formMUI

Video Renderer

repo

HyperFrames composition

A HyperFrames-authored video composition with TTS narration and transcript-driven highlights — the kind of asset I now use for product walkthroughs and launch reels.

HyperFramesTTSHTML compositions
▸ Local archive — ~20 repos across coding agents, voice, dev infra, and client SaaS. A subset shown.
05 / Stack & Principles

The tools
I trust.

~/manjot/stack.sh
utf-8
stack --list --by-domain
◆ AI · Agents
MCPs · Anthropic Claude · OpenAI · LangGraph · Vercel AI SDK · pgvector · Voice AI · RAG pipelines
◆ Frontend
React · Next.js · Vue · Angular · Remix · TypeScript
◆ Backend
Node.js · NestJS · Express · Spring Boot · Python (Django · Flask) · GraphQL · REST
◆ Databases
PostgreSQL · MySQL · MongoDB · CosmosDB · MariaDB · Teradata
◆ Cloud · Serverless
AWS Lambda · Azure Functions · Firebase · GCP · Vercel
◆ DevOps · CI/CD
GitHub Actions · Jenkins · Docker · Kubernetes · Azure DevOps
// Operating principles
01
Architecture before frameworks. Boring tech for boring problems.
02
Ship before perfect — feedback compounds, polish is cheap.
03
Strip AI-sounding patterns from every artifact.
04
Measure success rates, not vanity metrics.
Education
Lovely Professional University
M.Sc. Information Technology · 2019–20
Thapar University
B.E. Computer Science · 2014–18
// End transmission

Let's build
something real.

Three doors. Pick the one that fits. I reply within 24 hours, in plain English, and I'll tell you upfront if I'm the wrong fit.

01
Build with me

Client engagements

Agentic AI systems, voice agents, SaaS platforms, AI integrations on top of existing stacks. Best for funded teams who need senior architecture and want to ship in weeks, not quarters.

Send a brief →
02
Advisory

Fractional CTO · sparring

Architecture reviews, AI roadmap design, technical due diligence, hiring loops. Limited slots, paid by retainer or per engagement. Useful for founders making early stack bets.

Book a call →
03
Hello

Just say hi

Building something interesting? Want to argue about agent architectures, voice AI, or Indian quant infrastructure? My inbox is open. No pitch, no agenda required.

Drop a line →
— or reach out directly —
hi@msingh.co.in
+91 · 98034 32657