Skip to content
BackNov 7
Muhammad Awais A.
Upwork badge

Muhammad Awais A.

AI Engineer | Production RAG, AI Agents & Automation | Python, AWS

I build AI systems that hold up in production: RAG assistants that answer from your own documents with sources, AI agents that take real actions, and the Python backends and automations behind them. Most AI projects don't fail at the demo. They fail at the parts after it: messy documents, permissions, wrong answers nobody can debug, costs that grow with every user. That's the part I focus on. What I've built recently: • A multi-tenant RAG / GraphRAG platform (Qdrant + Neo4j) with hybrid retrieval, reranking, tenant-safe file access and source-grounded chat, for a Swiss AI-automation startup. • Knowledge ingestion from Google Drive, OneDrive, Dropbox and Notion: OAuth connections, scheduled sync, permission metadata, skipping unchanged files. • AI agent features on an Activepieces-based automation platform: templates, sandbox test mode, persistent memory, run-history explainability and streaming responses, with OpenAI, Anthropic and other LLM providers. • An AI-assisted accounting automation product (FastAPI + React/TypeScript): invoice and bill extraction, VAT checks, bank statement matching and a two-way Zoho Books sync. • For a repeat Upwork client: a Python RAG backend on AWS Lambda with autoscaling, Docker and OpenAI, across three back-to-back 5-star contracts. What I can do for you: • RAG / knowledge assistants: ingestion, chunking, hybrid search, reranking, evaluation, citations • AI agents and LLM features: tool calling, memory, guardrails, streaming UIs • Workflow automation: n8n, Activepieces, webhooks, OAuth integrations, CRM and accounting APIs • Backends: FastAPI / Node.js, PostgreSQL, vector databases, AWS (Lambda, EC2, DynamoDB, S3), Docker How I work: I start by finding the riskiest part of the project and proving it works in a small first milestone, so you see real results before committing to the full build. Clear written updates, clean code your team can maintain, and I overlap with both US and European business hours. "great python coder. aws knowledgeable. pleasure to work with. will invite for next project" (Upwork client, 5.0) If you have an AI or automation project, send me a message with what you're trying to get done. I'll tell you how I'd approach it and what the first milestone would be.

🇵🇰Multan, Pakistan$25/hr100% JSS5.00 (4)Since Sep 2022

Earnings and projects

MRR
$462

Last 6 months ÷ 6

World#81,211 / 412k
🇵🇰Country#10,196 / 52.9k
Revenue, last 30 days
$477

Last 6 months: $2.8k

Total earnings
$3.6k

Billed in all: $3.6k

World#282,453 / 412k
🇵🇰Country#33,169 / 52.9k
Per project
$516

$1.4k in the last 6 months

World#179,812 / 412k
🇵🇰Country#18,831 / 52.9k
Projects, last 6 months
2
100% fixed0% hourly
World#88,918 / 412k
🇵🇰Country#17,051 / 52.9k
Projects in all
7

Across 10 contracts

60% fixed40% hourly
World#236,188 / 412k
🇵🇰Country#38,061 / 52.9k

Earnings, last 30 days

$3.6k+$477 earned in the last 30 daysBilled in all, before Upwork's fee · drag across the chart to see what was earned in any stretch
$3,000$3,250$3,500$3,750No dataOct 8Oct 23Nov 7

Read once a week before Nov 5, 2026, daily since. Dotted: longer than usual between two readings. Hatched: over two weeks without one.

Skill rankings: Top 16% in AWS Lambda

Ranked by MRR among everyone on Upwork with the skill.

  1. #31Neo4j153 freelancers · 🇵🇰 #3Top 20%
  2. #320Vector Database1,142 freelancers · 🇵🇰 #110Top 28%
  3. #465AWS Lambda2,947 freelancers · 🇵🇰 #105Top 16%
  4. #709Chatbot Development3,106 freelancers · 🇵🇰 #275Top 23%
  5. #783Large Language Model2,726 freelancers · 🇵🇰 #245Top 29%
  6. #1,039LangChain3,821 freelancers · 🇵🇰 #423Top 27%
  7. #1,122Retrieval Augmented Generation3,618 freelancers · 🇵🇰 #417Top 31%
  8. #1,193FastAPI4,337 freelancers · 🇵🇰 #385Top 28%