I’m looking for a remote AI engineer role, full-time or contract, building LLM agents, RAG and evaluation systems that must be reliable and measurable. I work best on Python backends (FastAPI, PostgreSQL, AWS) where deterministic code checks model output. Based in Pakistan (UTC+5), available immediately, 40 hours a week, flexible on time-zone overlap.

Numan Farukh
@numanfarukh
AI Engineer with 8+ years in Python backends. Built FreightLens (LangGraph, Postgres) and RAG support agents; PodLead on Google Gemini agent data.
What I'm looking for
I’m an AI engineer with 8+ years of production software experience, mostly Python backends (FastAPI, Django), PostgreSQL and AWS, for clients in the US, Germany, Canada and Serbia. For the last two years I’ve focused on LLM agents, RAG and evaluation.
At NeuralFlex, I lead a team of about 12 ML engineers producing agentic tool-use training data for Google Gemini. I design scenarios with checkable expected outcomes and co-built the automated reviewer that quality-gates scenarios and their implementations before human sign-off.
I built FreightLens, a multi-tenant freight document intelligence SaaS that extracts and reconciles charges from carrier quotes and invoices. The LLM only reads; deterministic code validates the results. A LangGraph workflow adds repair loops and checkpointed human review, with Postgres row-level security, a CI quality gate and Terraform deployment to AWS. On 90 synthetic documents, charge-extraction F1 rose from 78.5% to 100%.
I also built TrailPeak, a RAG support agent with source citations, semantic caching, an evaluation harness with regression detection, and deployment on ECS Fargate.
I’m looking for a remote AI engineer role, full-time or contract, building LLM systems that need to be reliable and measurable. Available immediately and flexible on time-zone overlap.
Experience
Work history, roles, and key accomplishments
AI Engineer Projects
Independent projects
Aug 2026 - Present (3 months)
Built two production-style LLM systems end to end.
FreightLens: multi-tenant document intelligence SaaS (LangGraph, Postgres row-level security, CI quality gate, Terraform on AWS).
TrailPeak: RAG support agent with evaluation harness, semantic cache and ECS Fargate deployment.
PodLead, LLM Agent Training Data (Google Gemini)
NeuralFlex
Nov 2024 - Present (2 years)
Lead a team of about 12 ML engineers producing agentic tool-use training data for Google Gemini, including designing scenarios and co-building an automated reviewer.
Software Engineer (contract), InfoBud AI lead platform
100Days
Aug 2025 - Jan 2026 (5 months)
Built the full stack of InfoBud, a proof-of-concept lead-enrichment and campaign-matching platform, including async background workers and real-time progress streaming.
ML Engineer, LLM Agent Training Data (Google Gemini)
NeuralFlex
May 2024 - Nov 2024 (6 months)
Designed agentic scenarios and their Colab implementations for the same programme; promoted to PodLead after six months.
Full Stack Developer
100Days
May 2022 - Feb 2024 (1 year 9 months)
Built core features of a German cash-desk SaaS in a team of 6, including onboarding, chat, notifications and sharing, and integrated the Fiskaly API for fiscal compliance.
Senior Software Engineer
N3 Digital
Nov 2021 - Jan 2022 (2 months)
Built several client sites from scratch with payment-gateway integration and JWT authentication.
Full Stack Engineer
Loud Now Agency
Oct 2021 - Nov 2021 (1 month)
Implemented features and cleanups in a PHP/Yii and Next.js codebase, in a team of 10.
Senior Software Engineer
Batchbud
May 2021 - Oct 2021 (5 months)
Developed features for app.batchbud.com, moved front end from Django templates to Vue, and integrated Google APIs.
Education
Degrees, certifications, and relevant coursework
Punjab University College of Information Technology (PUCIT)
Bachelor of Science, Information Technology
2016 - 2020
Grade: 3.90/4.00
Activities and societies: Batch topper; multiple scholarships
BSc in Information Technology from Punjab University College of Information Technology (PUCIT), Lahore, with a CGPA of 3.90/4.00 and recognition as batch topper.
Tech stack
Software and tools used professionally
Amazon CloudWatch
Amazon S3
GitHub
Kubernetes
Docker Compose
minikube
GitHub Actions
PySpark
PostgreSQL
Django
Laravel
Next.js
Tailwind CSS
Nuxt.js
OpenCV
Redis
Terraform
React
Vue.js
JavaScript
Python
PHP
Locust
Celery
Django REST framework
FastAPI
SQLAlchemy
GraphQL
AWS Lambda
Deepgram
Amazon RDS
TypeScript
pytest
Docker
Amazon Elastic Container Service
Google Colab
LangChain
ChromaDB
Pydantic
OpenAI API
Groq
Vite
LangGraph
LangSmith
Firecrawl
Pipecat
Telnyx
AWS
Availability
Location
Authorized to work in
Portfolio
fanuman.github.ioSalary expectations
Social media
Job categories
Interested in hiring Numan?
You can contact Numan and 90k+ other talented remote workers on Himalayas.
Message NumanGet matched with your dream remote job
Sign up now and join over 250,000+ remote workers who receive personalized job alerts, curated job matches, and more for free!
