Bountystash is a technology based on information theory, risk analysis, and market data.

• Designed production knowledge-base pipelines that extract structured facts from unstructured documents and assemble retrieval context for downstream agents
• Built modular agent workflows with tool calling and Model Context Protocol so autonomous routines hit company APIs under explicit permissions
• Shipped evaluation harnesses that measure answer quality, hallucination rate, retrieval relevance, and tool-call reliability on every CI cycle
• Implemented Python and Go data pipelines that join relational sources with embedding indexes for low-latency context assembly
• Investigated production failures with traces and datasets, then iterated ranking and prompting against measured regressions
• Built risk-adjusted PPA and weather-derivative models for energy traders (in-house risk manager)

O3 Energy has been rated as one of the top solar developers/contractors in the US.

• Owned commercial and technical steering of a 100M+ USD / 10 MW solar portfolio, including provenance of operational and financial data
• Built dynamic models for entity-level risk, scenario analysis, and temporal performance of energy assets
• Shipped an internal mobile capture app that resolved project entities and fed a structured opportunity pipeline
• Designed a risk system that scored portfolio exposures and produced auditable traces for banks and transaction advisors
• Documented and managed software and IT admin for the company and subsidiaries

Smart Energy 360 is a top energy provider in the region.

• Ran a 100M+ USD project pipeline and translated client requirements into technical solution designs
• Advised on cybersecurity and risk tooling for energy operators
• Built sales and support processes that connected unstructured client intake to structured delivery plans

A world-renowned research institute in Germany and a foremost in solar technologies.

• Wrote C++ interfaces to scientific measurement systems and Python analysis for 3D geometry and error traces
• Verified geometric calculations in Mathematica and scored measurement-system failures in a research setting
• Conducted research on heliostat imperfections using interferometry

Halliburton is one of the top oil and gas service companies, Bakken Shale, Williston ND.

• Led two field crews on 3M+ USD/week operations with on-site equipment and chemical troubleshooting
• Shipped Excel engineering tools that replaced ad-hoc calculation and improved operational feedback loops
• Worked 12-hour field shifts; implemented new engineering spreadsheets

Risk Calculation Engine (Swirl Engine)

A polyglot risk calculation engine implementing Green's Theorem (Shoelace Formula) for stochastic shape analysis.

• Architected a Green's Theorem engine for stochastic polygon area calculation as a stress test of numerical pipelines
• Implemented a Go shape fuzzer generating provenance-tagged synthetic geometry
• Automated the data path with Nushell and Awk; containerized with Nix Flakes for reproducible Go builds

Master's Thesis

Artificial Intelligence applied to weather prediction

• Developed a time-cross-correlation method for wind-velocity data with a measured +70% accuracy lift
• Trained neural nets on long-horizon weather series to improve grid economics of intermittent solar and wind

Postgraduate Renewable Energy

Post-graduate research in renewable energy

• Analyzed hydrogen production and storage projects using RETScreen
• Post-graduate research in renewable energy
• Solar, Wind, Geothermal, and Bioenergy modules completed under the UNU programme