Conversational Analytics Agent
A data-science agent with 70+ natural-language tools.
A data-science agent with 70+ natural-language tools.
A Canvas automation tool for chaining workers into executable pipelines.
Spin a 3D globe and find live matches on the cities where they are played.
Grew a crash-simulation workbench agent to about 100 tools.
A BabylonJS viewer for 150+ MB automotive crash models.
A charting platform with 40+ chart types and ML model comparison.
Turns simulation results into presentations, PDFs and Word documents.
Shows who is using the compute cluster, when, and how hard.
Reproduces, root-causes, drafts fixes and screenshot-verifies.
Front end for 1,800+ journal sites, including top university titles.
Describe a decision you are stuck on; get branching alternate timelines.
Canvas game engines written from scratch, still playable.
Shipped a conversational data-science agent and scaled the 3D workbench agent to roughly 100 tools. Wrote the agent system templates and prompt contracts plus the results renderer. Set frontend architecture across a suite of automotive CAE web apps, took full ownership of the BabylonJS crash-simulation viewer, and mentor engineers.
Extended a BabylonJS/WebGL viewer for 150+ MB crash-simulation models. Hand-built the 16k+ line Canvas node editor behind the workflow builder, and built a Backbone and D3 v5 platform with 40+ chart types.
Graphics and interactive systems.
Front end for 1,800+ websites including top university journals, with cross-browser XHTML, CSS and jQuery.