📍 Lahore, Pakistan · Remote · Available for US / EU Hours

Backend engineer with 4+ years shipping production Python systems — Django, FastAPI, Flask — and the AI agent infrastructure that sits on top of them.

I own products end-to-end: designing data models, building async APIs, wiring up background workers, and deploying to AWS/GCP. I specialize in integrating LLMs and RAG pipelines into real SaaS products, including a B2B sales engine that lifted reply rates by 17% and a financial data platform that automated unstructured-PDF extraction for asset managers.

Equally effective as a solo first engineer or as the senior IC who tackles the hardest backend problems.

Python (expert) · Django · Django REST Framework · FastAPI · Flask · async/await · Celery · Redis · WebSockets · REST & GraphQL · OAuth2 / JWT · Multi-tenant architecture

RAG Architecture · LangChain · LlamaIndex · OpenAI API · Llama Fine-tuning (Alpaca, LoRA) · RLHF · Agentic Workflows · MCP & A2A Protocols · Prompt Engineering · Pinecone · Embedding Pipelines

PostgreSQL · MongoDB · Redis · Schema Design · Query Optimization · Alembic / Django Migrations · Data Vault 2.0 Modelling

AWS (ECS, EC2, Lambda, S3, RDS, Secrets Manager) · Terraform · Google Cloud · Docker · GitHub Actions CI/CD · Nginx · Gunicorn · Monitoring & Logging

React · Vite.js · JavaScript / TypeScript — can ship and maintain frontend independently

Twilio · WhatsApp Business API · Stripe · Jira SDK · Microsoft Graph / Outlook · Google Workspace · Puppeteer · Web Scraping

Nexusaims · Luxembourg (Remote) · Nov 2023 – Present

• Sole backend owner across 8 production AI products serving asset managers including Lemanik RIV Capital — from database schema through deployed REST APIs to scheduled background jobs.
• Built a multi-tenant Django + PostgreSQL platform with Celery + Redis for long-running LLM jobs, serving clean REST endpoints to both frontend and email-based clients.
• Architected a Unified Data Repository (UDR) that ingests unstructured PDFs and reconstructs structured financial data via a configuration-driven pipeline — eliminating ~3 weeks of manual data-entry per fund per quarter.
• Designed and shipped a production RAG system on LlamaIndex integrated with Outlook, answering investor and director queries directly from email with zero human handoff.
• Led requirements discovery with non-technical stakeholders, wrote technical specs, managed the roadmap, and onboarded contract developers — effectively operating as Fractional CTO.
• Automated compliance reporting and Due Diligence / RFP generation, reducing client operational headcount requirements by 2 FTEs.

smart-focus.ai · Remote · 2023 – 2024

• Built the entire backend solo — FastAPI, PostgreSQL, Redis queues — containerised on AWS ECS, then grew the team as revenue came in.
• Designed the core AI agent engine: researches B2B prospects across the web, synthesises context, and generates hyper-personalised cold emails, LinkedIn messages, and call scripts.
• Shipped a multi-tenant SaaS API with per-customer rate limiting, usage metering, webhook delivery, and OAuth integrations with clients' email stacks.
• Deployed into production at Engage3 across an 8-person SDR team — delivered a 17% lift in prospect reply rates, validated against CRM data.
• Owned everything below the UI layer: infrastructure, CI/CD, observability, customer onboarding, and technical sales calls.

Neoworlder · US Startup — Remote · Jul 2023 – Jan 2024

• Sole AI backend engineer; delivered a complete conversational AI stack from scratch for a metaverse product.
• Built a RAG pipeline using GPT-3.5, Pinecone, and Ada embeddings, wrapped behind a clean Python service layer with caching and retry logic.
• Fine-tuned Llama with Alpaca techniques for domain-specific agents, then folded RLHF into the feedback loop for continuous alignment.
• Designed a key-based conversational memory system maintaining context across long, multi-session user interactions.

Dexi.io (acquired by Engage3) · Remote · Sep 2022 – 2025

• Technical engineer on a web-scraping and retail-intelligence platform serving clients including Walmart — owned integrations, custom tooling, and scraping architecture.
• Built a Python/Jira-SDK internal tool that converts Customer Success tickets into client-ready emails, saving ~4 hours per week of manual CS work.
• Served as senior technical point of contact on data-pipeline design and large-scale extraction projects.

Self-Employed (clients incl. Edge By Ascencial) · Remote · 2019 – Aug 2022

• Delivered Python backend services and ML pipelines — sentiment analysis, aspect-based sentiment analysis, SVR regression, and BERT-based NLP classifiers — across 6+ client projects.
• Built a Digital Shelf Analytics system ingesting customer reviews through a multi-stage NLP pipeline, surfacing actionable brand insights via a Flask API.

Flask + Vite.js + MongoDB, Twilio BSP, Celery/Redis for scheduled messaging. Full tenant isolation, campaign scheduling, inbox UI, and CI/CD. Owned architecture and implementation end-to-end.

Configuration-driven MongoDB system ingesting unstructured PDFs/DOCX and extracting structured fund-management data via RAG + LLM. Designed the full pipeline: document parsing, attribute definitions, priority/Golden Record logic, async task queue, and audit trail.

RAG-powered agent managing personalised outreach and follow-up without human intervention. Integrated with Outlook via Microsoft Graph. Built entirely in Python with a FastAPI control plane.

Bachelor of Science in Computer Science · 2021 Punjab University College of Information Technology (PUCIT)

I'm open to senior backend and AI engineering roles — remote, US/EU hours.

📧 ahteshamulhaq002@gmail.com · 📱 +92 307 0148467