Intelligence.Applied.
I build at the intersection of models, systems,
and the real world.
Good AI doesn’t end
at the model.
It lives in the systems around it.
My work connects the pieces, from a local model
to an application people can actually use.
Local inference
Making capable models work within real hardware, memory, and latency constraints.
Agent orchestration
Connecting models, tools, and workflows into systems with a clear purpose.
RAG & knowledge systems
Bringing the right context to the right question, with sources you can inspect.
Evaluation & optimization
Understanding what works, where it fails, and which trade-offs matter.
Multimodal applications
Working across text, images, and other inputs to solve practical problems.
Forward-deployed engineering
Staying close to the problem, the people, and the path into production.
Real domains.
Real constraints.
Two areas of deepening focus.
One growing curiosity about the physical world.
Complex information.
Clearer decisions.
Exploring how language models, retrieval, and agents can make financial information more accessible, structured, and useful.
An area of focus, not a claim of deployed financial products.
Better systems.
For human work.
Interested in knowledge retrieval, multimodal information, and workflow tools that support healthcare professionals without obscuring their judgment.
An exploration area. No clinical tool or medical claims are presented here.
Intelligence meets
the physical world.
Following the connections between AI, robotics, and the mechanical, electrical, and structural systems that make real-world machines possible.
A learning direction, not an established robotics portfolio.
Ideas, becoming
systems.
The first case studies are taking shape.
This space will show the problem, the process,
and the evidence. Not just the polished result.
Building first.
Documenting next.
Projects will appear here as they are ready
to be shared.
Less black box.
More understanding.
A future home for hands-on explanations.
Explore the ideas behind the experiments
before the interactive versions arrive.
Trace a question
Follow a question through retrieval, context, and a grounded answer.
Follow an agent
Look inside a plan, its tool calls, and the checks between actions.
Explore the trade-offs
Compare model quality, latency, and memory on the same terms.
Blueprints today. Working experiments to follow. No live models are connected.
Papers. People.
A point of view.
Research worth unpacking. Ideas worth questioning.
My notes on what they mean beyond the headline.
More than
the headline.
A space for my readings of research papers, responses to experts, and original essays. Each connected to its source, with room for my own perspective.
Personally curated. No automated popularity rankings or invented news.
Notes from
the workbench.
Experiments, practical tutorials, and lessons
from making things work.
The first page
is still being written.
New field notes will appear here.
No filler posts. Just things worth sharing.
