Looking for a job

Noman Khan

@nomankhan13

AI engineer building production LLM systems, including a construction platform with AI email intake and Xero-integrated finance.

Pakistan

Message

On Brikline, I architected a multi-tenant construction management platform for a live Australian client, covering AI email intake, procurement, finance, and scheduling. I built a review queue for uncertain email classifications and human-reviewed AI-drafted replies.

I designed and shipped a voice AI patient registration system using Twilio and Groq, with a conversational state machine that collects, validates, and saves patient data. When a regional telephony restriction blocked live calls, I demonstrated the pipeline through simulated testing and deployed it to Render.

I also built a real-time AI tool-call gateway with policy enforcement and a three-layer NLP engine for PII redaction. For VibeExplain, I deployed a multi-tenant SaaS app using LangChain and ChromaDB, and created an algorithm to flag structural risk in code dependencies.

Experience

Work history, roles, and key accomplishments

Current

AI-Native Construction Management Platform

Current

Brikline

Jul 2026 - Present (3 months)

Architected a multi-tenant SaaS platform for a live Australian client, covering AI mailbox intake, lead-to-project conversion, procurement, finance, and scheduling. Built AI email classification with a confidence-based review queue and engineered a scheduling engine using Kahn's Topological Sort.

fastAPI PostgreSQL GROQ Celery Redis Next.js Xero Multi tenant architecture AI Email Classification Scheduling Kahn'S Algorithm Procurement Finance Cash Flow Forecasting

Current

VibeExplain

Current

VibeExplain

Apr 2026 - Present (6 months)

Deployed a multi-tenant SaaS app using LangChain, ChromaDB vector indexing, and intelligent LLM request routing. Created a Knowledge Concentration Score algorithm evaluating static code dependencies to flag structural risk.

Lang Chain RAG ChromaDB fastAPI PostgreSQL GROQ Multi tenant SaaS Vector Indexing LLM routing Knowledge Concentration Score

Current

Sovereign AI Gateway

Current

Sovereign AI Gateway

Feb 2026 - Present (8 months)

Built real-time AI tool-call middleware enforcing policy rulesets across ChatGPT, Claude, Copilot, and Gemini, with a 3-layer NLP engine achieving 39ms PII redaction latency. Validated 162/162 integration tests at 90.9 req/sec under 200 concurrent users.

Python fastAPI Docker XLM R Onnx HMAC NLP PII Masking AI Governance Middleware

Voice AI Patient Registration System

Jan 2026 - Dec 2026 (11 months)

Designed and shipped an end-to-end voice AI intake system using a telephony agent driven by an LLM conversational state machine. Replaced loose JSON parsing with strict schema validation and wrapped external API calls for graceful degradation.

fastAPI SQLite Twilio GROQ LLM Voice AI Telephony Schema Validation REST APIs