Custom AI Agents & Data Pipelines for Complex Workflows

The Data Bottleneck

Does this sound familiar:

Your raw data is ready, but your code is broken.
Deploying LLMs or vision models feels like a distraction from your core work.
You’re stuck wrestling with Python environments instead of analyzing results.

Work with me

I build secure, ready-to-run automation tools. From fine-tuning vision models (YOLO, nnU-Net) to deploying LLM data agents (Mistral), I turn your raw, messy data into actionable results so you can stop troubleshooting and start analyzing.

Who I help

For Business & Data Teams: Automate tedious data extraction, build custom LLM agents that process CSVs/XLS, and integrate cloud-vision APIs to scale your operations.
For Researchers & Labs: Stop wrestling with library dependencies. I build specialized image segmentation (nnU-Net, ResNet) and tailored analysis pipelines so you can focus on publishing.

Selected Projects

Here is a sample of recent pipelines I have built to extract high-value data from complex environments:

Autonomous Data-Processing Agents: Deployed a highly cost-effective, locally hosted LLM agent (Mistral Small) capable of ingesting raw datasets, writing and executing Python code, and generating analytical plots autonomously.
Cloud-Integrated Vision Pipelines: Developed a serverless Apify architecture connected to Runpod that ingests real estate URLs, extracts images, and uses a combination of YOLO and LLMs to automatically evaluate and describe property conditions at scale.
Precision Image Automation: Built an end-to-end pipeline combining Cellpose and FIJI to track complex nuclear structures, replacing hundreds of hours of manual annotation for a study currently under journal review.
High-Volume Signal Filtering (ResNet): Engineered a classification scaffolding to process thousands of raw Atomic Force Microscopy (AFM) curves, automatically isolating the <0.1% of valid signals from trash data.

The Technical Advantage

Flexible Deployment: I build where you need it. I can deploy pipelines strictly locally for maximum data privacy (academic/medical), or scale up via cloud infrastructure (Runpod, Apify) for high-volume processing.
Precision Engineering: Built for rigorous discovery and exact specifications, avoiding generic out-of-the-box compromises.
Ready-to-Run Delivery: Whether it is an isolated local pipeline or a cloud-based agent, I deliver working tools. No environments for you to configure.

Let’s discuss your challenge

Not sure if your specific problem can be engineered? Let’s see if I can help you!

Email me to book a Free 30-minute video consultation:

[email protected]

Ready to collaborate?

Choose the engagement model that fits your workflow:

Hire me via Email: Direct billing and custom contracts.
Hire via Malt: Project management via Malt platform.

Check my LinkedIn profile: