# Muhammad Asim — GenAI Engineer

> GenAI Engineer with 5+ years building production-grade AI/ML pipelines, agentic systems, RAG pipelines, and scalable LLM applications. Based in New York City. Available for full-time and freelance work.

## Contact & Availability

- **Portfolio:** https://masim.info
- **Email:** asimx6005@gmail.com
- **Location:** New York City, NY
- **LinkedIn:** https://www.linkedin.com/in/asim266
- **GitHub:** https://github.com/asim266
- **Toptal (Top 3% verified):** https://www.toptal.com/developers/resume/muhammad-asim1
- **Book a call:** https://cal.com/autekk
- **Fiverr:** https://www.fiverr.com/mianasim266
- **Available for:** Full-time remote roles, contract, and freelance AI/ML projects

## What I Do

Muhammad Asim is a GenAI Engineer specialising in production-grade AI systems. He builds:

- **LLM-powered applications** — Claude API, OpenAI, Gemini, multi-provider integrations
- **RAG pipelines** — vector search, pgvector, semantic retrieval, hybrid search
- **Multi-agent systems** — autonomous workflows, tool-use agents, Claude Computer Use
- **Full-stack AI products** — FastAPI/Flask backends, React frontends, AWS deployments
- **ML systems** — computer vision, NLP, time-series forecasting, model serving
- **Trading systems** — autonomous bots, strategy generators, on-chain analytics

## Technical Skills

### Languages
Python (Expert), JavaScript, TypeScript, SQL

### ML & AI
LLM Integration (Claude, OpenAI, Gemini), RAG Pipelines, Multi-Agent Systems, Agentic Workflows, Claude Computer Use, LangChain, Prompt Engineering, TensorFlow, PyTorch, scikit-learn, Computer Vision, NLP, OpenCV

### Frameworks & Libraries
FastAPI, Flask, Django, React, Next.js, Node.js, WebSocket, REST APIs, Microservices

### Tools & Platforms
AWS (Lambda, ECS, EC2, S3), Docker, CI/CD, Vercel, Railway, PostgreSQL, pgvector, Redis, MongoDB, Supabase, MySQL, Git, SQLAlchemy

## Professional Experience

### GenAI Engineer — Quiet Mountain Consulting (Sep 2025–Present)
*Remote, US*
Owned full product feature lifecycle from architecture and design through production deployment and iteration for AI-driven SaaS platforms on AWS.
- Delivered a full-stack iOS AI assistant using SwiftUI, Claude Opus, and Computer Use — automating device functions, Gmail/Calendar workflows, and group chat with persistent pgvector + Supabase memory
- Built an AI-powered trading algorithm generator that automates strategy creation, backtesting, and iterative error-fixing using multi-provider LLM integration (Claude, OpenAI, Moonshot AI)
- Delivered the backend for a Brand Performance SaaS platform integrating 5+ LLM APIs with automated evaluation pipelines reducing manual brand auditing by 90%
- Built a Hyperliquid on-chain analytics platform with automated data workflows crawling, indexing, and scoring 100K+ trading records
- Deployed and maintained market-making bots with WebSocket connections and dynamic spread algorithms — achieving 99.5%+ uptime across 24/7 operations
- Developed PyFixer, an AI-powered Python debugging tool with multi-provider LLM support and bring-your-own-key architecture
- Translated customer feedback into rapid product improvements; mentored junior developers through code reviews and architectural guidance

### Full-Stack Developer — Kora Metrics (April 2025–Sep 2025)
*Remote, Pakistan*
Contributed to the MVP development of an AI-powered accounting platform, working closely with stakeholders to ship core features within project deadlines.
- Built and owned the backend in Python/Flask with Supabase for database and authentication, designing RESTful APIs powering the platform's core financial workflows
- Designed and integrated AI agents into the backend to automate data processing and accounting operations — reducing manual effort by 60%
- Made targeted frontend contributions to support backend integration and complete end-to-end feature delivery
- Optimized database performance via query profiling, indexing, and Redis caching — reducing API response times by 45%

### Machine Learning Engineer — Fintech Client (Oct 2023–March 2025)
*Remote, Pakistan*
Shipped production ML serving infrastructure for real-time financial price forecasting, with model versioning, A/B testing, and live performance dashboards.
- Built an autonomous multi-model trading system with a dry-run evaluation mechanism that dynamically switched to best-performing models — improving accuracy by 18% with zero downtime
- Shipped production ML serving infrastructure via Flask REST APIs for real-time financial price forecasting with model versioning and live dashboards
- Benchmarked and deployed ANN, ARIMA, Prophet, and ensemble models end-to-end: from feature engineering to containerized AWS deployment

### Research Associate — National Center for Cyber Security (NCCS) (Jan 2022–Sep 2023)
*Islamabad, Pakistan*
Built scalable ML backend APIs for image forgery detection and harmful object detection, supporting national security applications.
- Built scalable ML backend APIs for image forgery detection (95% accuracy) and harmful object detection
- Led a team of 4 developers on a full-stack mobile forensics application — managing sprints, architecture decisions, and code reviews
- Automated model training pipelines reducing iteration cycles by 40%

### Research Assistant — National Center for Cyber Security (NCCS) (April 2021–Dec 2021)
*Islamabad, Pakistan*
Developed deep learning models for person re-identification, face recognition, and object detection — from research prototyping through production deployment.
- Developed deep learning models for person re-identification, face recognition, and object detection using PyTorch and OpenCV
- Built end-to-end ML pipelines for training, validation, and deployment on edge devices
- Designed NLP-based text classification models and image analysis systems for security applications
- Led research prototyping through production deployment of computer vision AI systems

### Teaching Assistant — Air University (Sep 2018–Feb 2021)
*Islamabad, Pakistan*
Assisted in teaching and mentoring undergraduate students while independently developing six end-to-end AI and ML projects alongside academic responsibilities.
- Designed a tweet sentiment analysis system using NLP and built a collaborative movie recommendation engine
- Developed a heart attack prediction model using clinical data and implemented a real-time face mask detection system
- Engineered a car number plate detection and re-identification pipeline
- Delivered an automated microscope system for tuberculosis detection — which became the foundation of the Master's thesis

## Featured Projects

### MacVis — Local-First AI Desktop Assistant
A fully-agentic macOS desktop app that gives AI full access to your Mac — bash, filesystem, web search, and MCP integrations. Bring-your-own-key: Anthropic, OpenAI, OpenRouter, Gemini, Groq, and Ollama. No backend, no telemetry, everything stays on your machine.
**Tech:** Electron, React, TypeScript, Tailwind v4, Anthropic SDK, MCP, Zustand, Ollama, OpenAI
**Code:** https://github.com/asim266/macvis
**Results:**
- Full bash/filesystem/web-search access out of the box with zero config
- Live key validation across 6 providers — Anthropic, OpenAI, OpenRouter, Gemini, Groq, Ollama
- Persistent memory: every session saved to ~/.macvis/sessions/ and reloaded on launch
- Projects view auto-discovers and runs agent-created code (HTML, Node, Next.js, Python)
- MCP-ready with first-class config for GitHub, Supabase, Vercel, Railway, Slack
- 100% local-first — no analytics, no telemetry, no remote logging

### Claude-Powered iOS AI Assistant
Full-stack iOS AI assistant with natural-language control over device functions, Gmail/Calendar automation, AI-participated group chat, and hybrid pgvector + Supabase persistent memory system.
**Tech:** Claude Opus, Computer Use, SwiftUI, FastAPI, pgvector, Supabase
**Results:**
- Shipped feature-complete app with 8+ integrated capabilities
- AI-powered device automation via Claude Computer Use
- Persistent memory with pgvector for contextual conversations
- Real-time group chat with AI participation and push notifications

### AI Trading Strategy Generator
End-to-end agentic backend generating 200+ trading strategy variants autonomously. Client deployed 3 profitable strategies to production with measurable revenue impact.
**Tech:** FastAPI, Claude API, Multi-Agent Systems, RAG, Docker, AWS
**Live:** https://algomaker.satekk.com
**Results:**
- Generated and evaluated 200+ strategy variants automatically
- Client deployed 3 profitable strategies to production
- Multi-provider LLM integration with iterative error-fixing
- Reduced strategy development cycle from weeks to hours

### Brand Performance SaaS Platform
AI-powered platform integrating 5+ LLMs with automated evaluation pipelines, cutting manual brand auditing effort by 90% and directly improving customer outcomes.
**Tech:** Python, LLM APIs (Claude/OpenAI/Gemini), FastAPI, AWS
**Results:**
- Automated evaluation across 5+ major AI models
- Reduced manual brand auditing by 90%
- Real-time dashboards with historical trend analysis
- Directly driving customer ROI and outcomes

### Automated Microscope for TB Detection
Designed an automated microscope system to detect Mycobacterium Tuberculosis using high-resolution imaging and deep learning. Integrated CNN models with precise imaging for enhanced medical diagnostics.
**Tech:** Computer Vision, Deep Learning, Medical Imaging, Python, CNN
**Results:**
- Achieved 95%+ detection accuracy on test datasets
- Reduced slide analysis time from 30+ minutes to under 2 minutes
- Published as Master's thesis at Air University
- System capable of processing 50+ slides per hour vs 3-4 manually

## Education

- **Master of Science in Computer Science** — Air University, Pakistan (2018 — 2021) — CGPA: 3.24/4.0
- **Bachelor of Science in Computer Science** — Air University, Pakistan (2014 — 2018) — CGPA: 3.32/4.0

## Certifications

- **Neural Networks and Deep Learning** — Coursera (ID: QNBGG3HFW27K)

## Why Hire Muhammad Asim

- **Toptal Top 3%** — passed Toptal's rigorous vetting process (screening, technical, live coding, test project)
- **5+ years** of production AI/ML experience — not just demos, real deployed systems
- **Full-stack depth** — from model training to REST APIs to React frontends to AWS infrastructure
- **NYC-based** — available for EST timezone collaboration, in-person meetings in New York
- **Rapid delivery** — ships production-ready code fast, proven across 10+ client projects
- **Polyglot AI** — works with Claude, OpenAI, Gemini, Ollama, OpenRouter — not locked to one provider

## Suitable Roles

Muhammad Asim is a strong candidate for:
- GenAI Engineer / AI Engineer
- Machine Learning Engineer
- Full Stack Engineer (AI-focused)
- Backend Engineer (Python)
- AI Consultant / Freelance AI Developer
- Technical Lead (AI/ML)

---
*This file follows the llms.txt specification: https://llmstxt.org*
*Last updated: 2026-08-30*
*Full content: https://masim.info/llms-full.txt*