Fasih Ullah SaleemAI Engineer
AI Engineer | LLMs | RAG | Python
I build LLM-powered applications, Retrieval-Augmented Generation (RAG), vector databases, prompt engineering and AI agent workflows, on a strong Python and backend foundation (FastAPI, REST APIs, databases). I have hands-on experience deploying and benchmarking open-source models locally with LM Studio and Ollama. Electrical Engineering background (CGPA 3.79/4.00); a fast learner comfortable owning AI features from experimentation to deployment in remote, international teams.
Open to remote and on-site roles worldwide · Based in Pakistan · Available for freelance, contract and hobby-project work
In the stack
Where an ai engineer sits in a connected product.
The solid layers are the ones this role owns: cloud & backend, applied ai. The faded ones are the rest of the system I also build, so nothing is designed in isolation.
- 04Cloud & backend
- 06Applied AI
Core competencies
What I bring to an ai engineer role.
- LLM Application Development
- Retrieval-Augmented Generation (RAG)
- Vector Databases
- Prompt Engineering
- AI Agent Workflows
- Local Model Deployment
- AI-Assisted Development
- Version Control (Git)
Hire me for
Three things I can start on next week.
For a team, a startup, a research group or one person with an idea. Remote, worldwide, and I reply within a day.
- 01
A chatbot or assistant grounded in your own documents with RAG, on a local model if your data cannot leave the building.
- 02
AI agent workflows that automate a real process, with a FastAPI backend and honest evaluation of what the model gets wrong.
- 03
Benchmarking and deploying open-source LLMs on hardware you own, instead of paying per token.
Technical skills
Tools and technologies.
AI & Intelligent Systems
- Local LLM Deployment
- Retrieval-Augmented Generation (RAG)
- Vector Databases
- Prompt Engineering
- AI Agent Workflows
- Local Chatbot Development
- Ollama
- LM Studio
- ComfyUI
AI-Assisted Development
- Claude & Claude Code
- ChatGPT
- GitHub Copilot / OpenAI Codex
- Cursor
- AI Pair-Programming
- AI-Driven Debugging & Code Review
Backend & APIs
- Python
- FastAPI
- Django
- REST APIs
- PostgreSQL
- Redis
- Automation Scripts
Infrastructure
- Docker
- Nginx
- Linux (Ubuntu)
- GPU Workstation Builds
- VMware ESXi
- PyTest
Relevant projects
AI Engineer work, built and photographed on my own bench.
Each card opens the full story: what it does, how it was built and which layers of the stack I built myself.
Research & prototyping
A local-first AI assistant built on open-source LLMs running entirely on-device. I explored Retrieval-Augmented Generation and vector-database concepts to ground the model in a private knowledge base, then experimented with prompt engineering, conversational interfaces, and agent workflows, evaluating model quality, hardware requirements, and deployment trade-offs along the way.
Embedded + cloud pipeline
An IoT data pipeline that takes readings from embedded sensor nodes and publishes them over MQTT to a cloud database, where they feed into monitoring applications. The project demonstrates the full telemetry loop: device acquisition, lightweight pub/sub transport, persistence, and live visualization.
Lead: firmware, PCB and integration
An end-to-end fleet telematics system built on a custom STM32F427 board. The device reads GPS position and MPU6050 motion data in real time, detects accidents and rollovers, and streams everything over a SIM800L GSM link to a web dashboard. A full power subsystem handles automatic switching between the vehicle battery and an onboard backup, with charging and protection circuitry, so the unit stays online even if main power is cut.
Experience and education
Engineering work, in production, since 2024.
An application engineer on simulation-driven power-system and engineering software projects, with an Electrical Engineering degree and PEC registration underneath.
- June 2025 to presentPakistan · remote-friendly
Application Engineer
- Design, evaluate and implement engineering software solutions: system architecture, workflow optimisation and technical feasibility.
- Build and extend engineering applications in Python and FastAPI, turning technical requirements into working backend features and automation.
- Develop automation scripts, scheduled tasks and database-backed workflows; work on simulation-driven power-system projects and design validation.
- Deploy, configure and troubleshoot applications in Linux and virtualised (VMware ESXi) environments.
- Run technical investigations, feasibility studies and root-cause analysis across software, infrastructure and engineering systems.
- Write technical documentation and design-review material for a multidisciplinary electrical, software and industrial team.
- June 2024 to June 2025Pakistan · remote-friendly
Engineering Intern
- Backend development, database tasks, automation, testing and software validation.
- Design reviews and UI/UX work in Figma.
- Engineering simulations, data analysis and technical documentation.
- 2021 to 2025Lahore, Pakistan
Bachelor of Science in Electrical Engineering
University of Engineering and Technology, Lahore
- CGPA: 3.79 / 4.00
- Registered Engineer, Pakistan Engineering Council (PEC)
Questions
What people ask before hiring an ai engineer.
What is RAG and when does a business need it?
Retrieval-Augmented Generation looks up the relevant passages from your own documents in a vector database and hands them to the model with the question, so answers are grounded in your knowledge rather than the model’s memory. You need it whenever the answer has to be specific, current and yours: manuals, policies, product data, case files.
Can you run AI models locally instead of using a cloud API?
Yes. I deploy and benchmark open-source LLMs with Ollama and LM Studio on hardware I built myself, and I evaluate model quality, VRAM requirements and throughput before recommending local versus hosted. Local is the right call when data privacy, cost per query or offline operation matter.
Do you build the backend around the AI as well?
Always, a model on its own is a demo. I build the FastAPI service, the ingestion and vector-indexing pipeline, the database and the deployment, so the assistant is a product, not a notebook.
Are you available for remote AI engineering roles or freelance AI projects?
Open to remote and on-site roles worldwide, and freelance AI projects for clients anywhere, including a first RAG prototype for a small team that wants to see it work on their own data before committing.
Also hire me as
The same engineer, seen from other angles.
One person, nine job titles. Each page argues the case from that field's side, with the projects that prove it.
Contact
Let's build something that works.
Hiring for a team, or need a product, prototype or hobby project built end to end? Either way it starts with one email. I reply within a day.
- Phone / WhatsApp
- +92 320 9450014
- Based in
- Pakistan · open to remote and on-site roles worldwide
