Research Engineer who ships.
I bridge the gap between academic research and production systems. I hold an MS CS (Georgia Tech) with published research at NeurIPS/NAACL, but my focus is on building things that work in the real world and understanding why they break.
My writing on human expertise and AI hit #1 on r/ArtificialIntelligence (>200k views). My essay on my own resilience was shared by Jeff Dean.
What I build:
End-to-End Local Voice AI: A privacy-first voice assistant running locally using Whisper + Ollama + Mistral. Full-stack integration of audio and LLM modalities on consumer hardware.
Automation from the Inside: I currently work retail while rebuilding my career. I wrote a technical analysis of why store automation fails, examining sim-to-real drift, impedance mismatch, and latent provenance. Not from a lab. From the floor.
Research training under prominent AI researcher. I know how to go from paper to experiment design.
What I'm looking for:
A role in ML Engineering, Applied Research, or LLM Evaluation where the job is to take cutting-edge systems and make them robust in the real world. I'm interested in the seams, where models meet reality and break. Whether that's building or stress-testing, I want to be close to the failure modes.
Location: Chattanooga, TN, USA
Remote: Yes
Willing to relocate: Open
Technologies: Python, PyTorch, Hugging Face, OpenAI APIs, LangChain, SQL, Matillion ETL, basic Gradio/Streamlit. Picking up React, TypeScript, and full-stack deployment skills.
Résumé/CV: https://www.linkedin.com/in/tahamerghani
Github:https://github.com/taha-y-merghani
Email: taha.y.merghani@gmail.com
Worked across NLP research, data engineering, and product prototyping. I’ve shipped projects solo and within early-stage teams—recently built a full LLM-based chatbot with voice input and a Gradio UI on old hardware. I’m comfortable navigating ambiguity, learning fast, and moving from idea to working demo.
I’ve worked in environments where nothing was set up—no infra, no roadmap, just a hunch that AI could help. So I got good at asking the right questions, tracing things to first principles, and building what mattered. If you’re looking for someone who thinks clearly, learns fast, and isn’t afraid to own the boring parts too, I’d love to connect.
My background is in NLP, research, and startups. I joined a power company where I saw a clear opportunity to use AI for automating equipment inspection from drone images.
But the environment made it hard to move fast. The systems were outdated, and there wasn’t much support for building AI tools in-house. That experience made me realize I needed to grow beyond the modeling layer. There were things I wanted to build, but I didn’t yet have the full skill set to do it on my own.
So I’ve been learning full stack development. I had built a small chatbot app before, but this time I’m applying what I’m learning toward a focused MVP for the inspection work. It’s been a practical way to connect what I know with what I want to make real.
Funny, I actually have a very similar story, where the plan was also to use drones/AI for inspection of power equipment. For the same reasons as you I quit to work on my own projects, but I discarded the drone project and went another way. Best of luck!
I grew up in Sudan. Learned to memorize scripture before I ever touched a keyboard. Came to the US, made it into Georgia Tech, and almost fell apart trying to stay there.
This piece isn’t a teardown of academia or a playbook for job hunting.
It’s a reflection. On what it means to move through systems that weren’t built for you, and how easy it is to get overlooked when your story doesn’t fit the frame.
It’s also about building. Not just code, but clarity. Not just skill, but self-definition.
If you’ve ever felt like the best parts of you didn’t show up on paper, or got filtered out before they were seen, this might resonate.
This started as a conversation with a friend who doesn’t play games.
I tried to explain why I still lose despite thousands of hours in Tekken 8.
The deeper I went, the more I saw it wasn’t just about games. It’s about reaction time, pattern recognition, psychology — and how you hit a wall even when you know what to do.
Would love to hear how others have experienced this in games, coding, or life.
Remote: Yes
Willing to relocate: Yes
Technologies: Python, PyTorch, Whisper/Ollama, Local LLMs, Research Synthesis
Résumé/CV: https://taha-y-merghani.github.io/Taha_Merghani_Resume.pdf
Email: taha.y.merghani@gmail.com
Research Engineer who ships. I bridge the gap between academic research and production systems. I hold an MS CS (Georgia Tech) with published research at NeurIPS/NAACL, but my focus is on building things that work in the real world and understanding why they break.
My writing on human expertise and AI hit #1 on r/ArtificialIntelligence (>200k views). My essay on my own resilience was shared by Jeff Dean.
What I build: End-to-End Local Voice AI: A privacy-first voice assistant running locally using Whisper + Ollama + Mistral. Full-stack integration of audio and LLM modalities on consumer hardware. Automation from the Inside: I currently work retail while rebuilding my career. I wrote a technical analysis of why store automation fails, examining sim-to-real drift, impedance mismatch, and latent provenance. Not from a lab. From the floor. Research training under prominent AI researcher. I know how to go from paper to experiment design.
What I'm looking for: A role in ML Engineering, Applied Research, or LLM Evaluation where the job is to take cutting-edge systems and make them robust in the real world. I'm interested in the seams, where models meet reality and break. Whether that's building or stress-testing, I want to be close to the failure modes.