Portfolio / Berkeley, California
Systems thinker. Builder.
I build forecasting tools, public air-quality data platforms, and AI assistants. My experience spans enterprise data strategy, production software, and the decisions that connect them.
01 / Independent projects
Player forecasts and multi-week squad planning for Fantasy Premier League, built around a separately testable prediction model and optimizer.
Live product. Open model, weights, and solver. Evaluated across 24 clean gameweeks with three seeds.
An open air-quality platform that turns fragmented government, community-sensor, and aggregator measurements into traceable public data.
73 cities and towns, 334+ stations, and 10M+ collected measurements. Public-site snapshot: September 8, 2026.
StudentWhat's due this week?
Causal Inference / Midterm 1
An AI study assistant connected to Berkeley course materials, deadlines, and grades. Python computes the facts; the model explains them.
12 read-only tools. 5 prompts. No write access. Runs on the student's laptop with their own bCourses token.
02 / Enterprise work
Data & AI Solutions Architect
Built an Azure OpenAI observability service and rolled out tracing across 10 .NET applications. AI-assisted analysis reduced Azure infrastructure cost by 15%; a new delivery workflow paired generated code with independent verification across 12 repositories.
Data, Analytics and AI Lead; Manager
Led the data workstream for a $176.4M fintech transaction across 10+ source systems, location intelligence for 320+ retail sites, and predictive maintenance that reduced unplanned downtime by 25%.
Data Scientist
Built six production ML models for telecom, automotive, and marketplace clients in Southeast Asia, contributing to a 30% improvement in marketing ROI and an 11% increase in conversion.
Guidant / Architecture & validation design
At Guidant Financial, I designed an AI-assisted validation layer for retirement-plan data, with deterministic controls and an in-tenant AI boundary defined before implementation.
Feb - Jun 2026
Data & AI strategy consultant
Authored the Data & AI Development Plan, Operations Manual, architecture papers, and executive presentations for an AI-native city.
Defined data ownership, sharing, and AI governance frameworks, alongside technical concepts for digital twins, geospatial data, knowledge graphs, and semantic exchange. Strategy and design work, not a deployed city platform.
2026
Knowledge-graph architecture
A decision guide for adding a semantic layer to a data marketplace, from competency questions and ontology governance to validation, access control, and consumption.
Compared property graphs, RDF stores, and hybrid patterns; defined a phased path from a minimal ontology to SHACL-validated production use.
03 / Experiments & research
Prompt linting in the browser, with a separately evaluated fine-tuned model for deeper analysis.
Write-up 05A voice tutor that asks follow-up questions, with local speech processing and conversation-level evaluation.
Write-up 06An intraday ranking model tested across four walk-forward quarters, with explicit leakage and promotion checks.
Write-up04 / Working principles
Compute dates, grades, and permissions in code. Give the model facts to explain.
OpenMind / deterministic coreTest prediction and optimization separately. A better forecast does not automatically make a better plan.
SmartPlay / prediction and optimizationPreserve the original reading, its unit, and the checks it passed.
AirData / immutable source dataSet promotion criteria before an experiment. Keep changes that survive each walk-forward quarter.
Four Sigma / promotion gatesA good reply can still make a poor tutor. Test repetition, answer leakage, and progress across turns.
Loqui Loop / thread evaluation05 / Background
I grew up in Kazakhstan and have worked across Astana, Seattle, Singapore, Almaty, and Berkeley. Outside the work, I have finished the Almaty Half Marathon nine times, including a sub-two-hour run.
Master of Information Management and Systems
2025 - 2027 (expected)Artificial Intelligence Programme - 89%, top five in cohort
2023BA, Economics and Business
Chevening Scholar, 2025 / Bolashak Scholar
Get in touch