Shipping AI products for clients across US · UAE · Germany · Saudi Arabia

Mohammad
Adil Sheikh

$ AI Engineer — ▌

I build production multi-agent systems, voice AI pipelines, and GTM automation that ship — agentic research systems that compress 2 days of SDR work into 2 minutes, and proposal pipelines that generate enterprise-grade documents in under 60 seconds.

View Projects GitHub Email Me
0Years building AI systems
0Voice calls automated / month
0AI models orchestrated in one pipeline
0To generate a 22-page proposal
0SDR research time eliminated

01. Featured Projects

Production systems, not demos — these run for real businesses today.

code ↗

ProposalForge

Agentic GTM Intelligence Platform

End-to-end agentic pipeline for B2B consulting: scrapes prospect websites, runs 5 parallel LLM analyses (industry, pain points, competitors, decision-makers, ICP fit), and auto-generates Big-4-grade 16–22 page PDF proposals in <60s — with Gantt charts, budgets, and team bios.

  • Next.js
  • TypeScript
  • FastAPI
  • PostgreSQL
  • Puppeteer
  • Claude
  • LangGraph

Multi-Agent Prospect Research

5-Agent GTM Research System

5 specialized agents — researcher, enricher, scorer, competitor analyst, outreach drafter — collaborate via shared context to research accounts and draft personalized outreach. 2 days of SDR work → ~2 minutes. Pulls intent signals from Reddit, X, HN, and LinkedIn.

  • LangChain
  • LangGraph
  • Claude
  • Python
  • Scraping APIs

mediaOS

AI Video Ad Production Pipeline

Production SaaS for DTC brands: competitor video breakdown → RAG-powered scriptwriting → multi-model scene generation across 11 video models (Kling, Sora, Veo, Seedance) + 5 voice models → one-click Meta Ads deployment with automated 7-day ROAS tracking.

  • FastAPI
  • React
  • PostgreSQL
  • Celery
  • 34 AI Models
  • Meta Ads API
code ↗

Multi-Agent Orchestration Platform

Autonomous Agent System

Production-grade autonomous agent platform: 14 skills, 5 specialized agents (researcher, coder, sales), multi-channel control via Web, Telegram, and webhooks, plus cron-scheduled autonomous jobs with failure-recovery loops.

  • Next.js
  • LangChain
  • LangGraph
  • Docker
  • Telegram
  • GitHub Actions

Voice AI Batch-Calling System

10K+ Calls / Month in Production

Automated EMI reminder system with bilingual Hindi/English agents that switch language in real time. SIP trunk integration, AI transcript analysis for call categorization and lead scoring — drove a 40% improvement in customer response rates.

  • LiveKit
  • Gemini
  • Deepgram
  • Sarvam
  • MongoDB
  • FastAPI
code ↗

AI Avatar Speaking Coach

Real-Time Lip-Synced Video Avatar

Real-time speaking coach with a lip-synced video avatar. Pipeline: Deepgram STT → Gemini LLM → ElevenLabs TTS with preemptive generation streaming for sub-300ms latency and BVC noise cancellation.

  • LiveKit
  • Tavus
  • Gemini 2.5 Flash
  • ElevenLabs
  • Silero VAD

02. Experience

Mar 2026 — Present

AI Full Stack Manager · Eaglytics Co (Remote)

  • Architecting GTM AI agents — built ProposalForge, generating enterprise proposals in under 60 seconds from a prospect URL.
  • Built the 5-agent prospect research system collapsing 2 days of SDR work into ~2 minutes.
  • Designed mediaOS — 11 video models, 5 voice models, and image generators unified in one ad-production pipeline for DTC brands.
  • Instrumented agents with structured logging, monitoring, and failure-recovery loops.

Oct 2025 — Mar 2026

AI Engineer · Virtual Galaxy Infotech Ltd, Nagpur

  • Built production voice AI batch-calling handling 10,000+ calls/month with LiveKit, Gemini, Deepgram, and SIP trunking.
  • Bilingual Hindi/English voice agents with human-like flow — +40% customer response rates.
  • Automated transcript analysis with Gemini for categorization, lead scoring, and follow-up scheduling.

Jan 2025 — Oct 2025

Freelance AI Automation Developer · US, Germany, Saudi Arabia

  • Delivered end-to-end voice agents, CRM-integrated automations, and multi-agent orchestration for international clients.
  • Shipped an AI Avatar Speaking Coach with sub-300ms latency (Tavus + LiveKit + Gemini + ElevenLabs).
  • Built Upwork→Slack lead bots and N8N workflows for outbound automation.

Sep 2022 — Dec 2024

Software Engineer · HCL Technologies, Nagpur

  • Built responsive web apps with React, Redux, and Node/Express REST APIs.
  • 30% faster load times via code splitting and lazy loading; CI/CD with Docker and GitHub Actions.

03. Technical Arsenal

// AI & Agents

  • LangChain
  • LangGraph
  • CrewAI
  • MCP
  • RAG
  • Claude
  • Gemini
  • OpenAI

// Voice AI

  • LiveKit
  • Deepgram
  • ElevenLabs
  • Sarvam
  • Silero VAD
  • SIP Trunking

// Video & Creative AI

  • Kling
  • Sora
  • Veo
  • Seedance
  • Tavus
  • fal.ai

// Backend

  • Python
  • FastAPI
  • Node.js
  • Express
  • Celery
  • REST APIs

// Frontend

  • React
  • Next.js
  • TypeScript
  • Vite

// Data & Infra

  • PostgreSQL
  • MongoDB
  • Redis
  • pgvector
  • Docker
  • GCP
Google Cloud Professional Cloud ArchitectCertified
Google Cloud Professional Cloud Security EngineerCertified
Google Cloud SpeakerCommunity

04. Get In Touch

Open to AI engineering roles and consulting — multi-agent systems, voice AI, and GTM automation. If you're building something ambitious, let's talk.

madilsheikh7@gmail.com GitHub LinkedIn

📍 Nagpur, India · working globally (remote)