I design production-ready systems and research-grade models that solve real problems across finance, agriculture, and local businesses — combining reproducible research with engineering pragmatism, from backtests and notebooks to secure APIs and deployed platforms.
I'm a software & data engineer with a Master's in Financial Engineering and a background building production systems and research projects. I've built inventory & invoicing platforms for SMEs, courier systems, and deep reinforcement learning agents for portfolio optimization.
I focus on reproducible research, clear engineering, and delivering measurable impact — preferring concrete metrics like Sharpe ratio, ROI, and latency over vague claims of success.
I combine rigorous research methods with engineering pragmatism — from reproducible notebooks and backtests to secure API design and production deployment. I document decisions, share code, and prefer clear metrics (Sharpe, ROI, latency) to evaluate outcomes.
Selected work spanning deep reinforcement learning for portfolio allocation, ESG-aware finance, and applied simulation for agriculture.
A PPO-based reinforcement learning agent for dynamic portfolio allocation, combined with a Hidden Markov Model for market-regime detection, evaluated through robust backtesting against standard benchmarks.
Integrating ESG scoring, sentiment analysis, and regulatory-risk forecasting into a climate-aware approach to financial risk modeling.
A simulation-based feasibility study combining agent-based modeling with NPV/ROI economic analysis and climate-aware AI modules for small-scale hydroponic farming.
Click through to GitHub for source, or open the research paper for the DRL portfolio work.
Personal website outlining skills, experience, projects and education.
Inventory management, SMS order updates, Excel/PDF exports, order tracking and QR codes.
React frontend, Spring Boot backend, and Flutter mobile apps for parcel routing and tracking.
PPO agent optimizing Sharpe ratio with HMM-based regime detection, validated through robust backtests against benchmarks.
Simulation-based feasibility study with NPV/ROI analysis and climate-aware AI modules.
ESG scoring, sentiment analysis, and regulatory risk forecasting for climate-aware portfolios.
Awarded first place for a research presentation at a national scientific conference, recognising excellence in academic research and innovation.
Built a Vue.js campus-platform prototype with booking, schedules, news, ratings, authenticated REST APIs, and validation logic; awarded Best Web Solution.
East Siberia State University of Technology and Management — currently in progress.
WorldQuant University
Copperbelt University
A running archive of the research and projects above, written up in more depth.
How a PPO agent and an HMM regime detector combine for dynamic portfolio allocation, and what the backtests show against standard benchmarks.
Read paper →Notes on combining sentiment analysis and ESG scoring with regulatory-risk forecasting for climate-aware portfolios.
View on GitHub →Simulation and economic modeling behind an AI-assisted feasibility study for small-scale hydroponic farming.
View on GitHub →More research notes and technical write-ups are planned for this section.
Research methodology, AI/ML strategy, and data pipelines for organisations, researchers, and data-driven projects.
Business model design, technology architecture, and AI adoption strategy for founders and teams building or scaling a venture.
Financial modelling, valuation, and portfolio or real-estate investment analysis for businesses and investors.
ESG and climate-risk analytics, financial risk modelling, and AI-informed approaches to sustainable investment decisions.
Interested in AI research, quantitative finance, data systems, or collaborative projects?