I turn operational inputs—floorplans, documents, video, and workflows—into usable AI products, decisions, reports, and tools.
01 / Selected work
A focused selection of projects where I turned manual, document-heavy, and plan-heavy operations into reliable internal products.
AI system for technical and compliance verification on floorplans and large construction documents.
Teams were manually reading 40+ page documents and running compliance checks by hand on floorplans.
Brickroad is a dApp where users can monetize their datasets.
AI pipeline that converts construction-site videos into compliance-report drafts.
Inspectors and engineers had to take notes on site, return to office, and write reports manually.
R&D project exploring practical techniques to detect and extract floorplan polygons with a reusable training and inference method.
Strong dependency on Kreo blocked extension to new plan classes like electrical and plumbing.
+ More projects
Confidentiality / Sensitive client details are anonymized. Project descriptions focus on workflows, architecture, and outcomes.
02 / Expertise
Core capabilities I use in long-term collaborations to ship business-ready systems.
03 / How I work
I integrate with small engineering teams (5–10 devs) and build incrementally with clear validation loops.
Frame the business workflow and target outcome
Break down features and dependency tree
Validate scope and priorities with the team
Build a functional MVP
Iterate feature by feature on real usage
Stabilize, scale, and transfer to internal operations
04 / Writing
I write about agent architecture, Claude Code workflows, AI-powered knowledge systems, and lessons from building with AI.
Medium article
A practical argument for focusing on harness design, feedback loops, tools, and evaluation instead of only swapping models.
Medium article
A breakdown of how architecture choices, context management, and cleaner agent boundaries reduced token consumption.
05 / About
I build AI systems that reduce operational time and make complex workflows easier to run.
I hold an engineering degree specialized in Data & Artificial Intelligence, with 7 years in software development, 3 years in Data Science, and 1.5 years as a freelancer.
At BTP Consultants, I work as an Expert Data Scientist and built end-to-end Computer Vision pipelines (collection, dataset, training, inference), RAG architectures for large documents, and APIs for floorplan interaction (polygon detection, measurements, standards verification).
I also built AI Agents interacting directly with floorplans, helping reduce verification time by 80%.
I’m known for fast adaptability across stacks, strong product sense, and a calm collaboration style focused on business outcomes.
06 / Contact
I’m open to long-term missions in AI Agents, Computer Vision, Data Science, and workflow automation.
In your first message, include your company name, business context, objective, and expected timeline.
