Danila
Kuznetsov

Quant ML / Applied AI Engineer focused on crypto markets, time-series ML and production-grade analytical systems. I build end-to-end research-to-production solutions: market data pipelines, feature engineering, leakage-safe validation, ML models, realtime WebSocket infrastructure, inference services, monitoring and analytical dashboards. My private Crypto Alpha Intelligence Platform combines quantitative ML, BTC market regime/direction models, futures & order-flow analytics, derivatives/options context, SQLite data infrastructure, backend services, LLM-assisted analysis and forward outcome monitoring. Strong in Python, SQL, data engineering, statistics, market analysis and backend development. I can take a quantitative or AI idea from raw data and research to a reliable working system. Open to remote contract, project-based and long-term work in Quant ML, Crypto, Trading Infrastructure, Applied AI and Market Data.

https://github.com/danila-quant-ml

Work experience

Independent / Private Project

September 2020 – Current time

Job title

Quant ML / Applied AI Engineer — Crypto Research Platform

Work experience & achievements

Built a private production-like crypto market intelligence platform covering the full research-to-production lifecycle. • Designed Quant ML models for BTC market regime and directional analysis. • Built leakage-safe feature engineering and walk-forward/pretest validation pipelines. • Developed model inference with frozen feature lists, thresholds, data-quality checks and signal logging. • Built realtime WebSocket market-data infrastructure with trades, BBO, CVD, delta volume and VWAP. • Added derivatives/options-flow and market microstructure analytics. • Built Python backend services, SQLite persistence, dashboards, monitoring and outcome review. • Integrated controlled LLM components for market context, news intelligence, risk flags and reporting. • Added automated workflows, tests, health checks and forward monitoring. The project demonstrates my ability to take quantitative research from raw data to a reliable live analytical system.

Education

Not specified

Crypto Alpha Intelligence Platform — Quant ML, Realtime Market Intelligence

End-to-end BTC market intelligence platform combining Quant ML, realtime market data, order flow, derivatives/options analytics, LLM intelligence and production monitoring. Built as a private research-to-production system: market data → ML inference → contextual analytics → logging → dashboard → outcome review. Not a trading bot or notebook demo. Full architecture, selected results and engineering case study are available on GitHub.

Quant ML for BTC — Models, Validation, Production Handoff

Custom Quant ML stack for BTC covering market regime, direction, macro expansion risk and post-04:05 statistical modeling. Selected results include 96% pooled precision for the Expansion STRONG tier, 0.8078 mean balanced accuracy for Direction, and 90.5% precision on the News + BTC final holdout. Built with strict temporal validation, leakage protection, frozen feature contracts and production model packaging. Full methodology and metrics on GitHub.

Realtime Crypto Market Data Infrastructure

Realtime BTC market-data infrastructure built from public exchange WebSockets to a live analytical dashboard. The system processes trades and BBO into 1s/5s/1m flow, CVD, delta and VWAP, persists data in SQLite, and securely streams live market context from VPS infrastructure to a local dashboard. A complete realtime data pipeline — not just an API integration. Architecture and implementation case study on GitHub.

Market Microstructure, Derivatives Intelligence

Research system for BTC market microstructure, options flow and price-level reactions. Combines Deribit options intelligence, Options Zones, pressure, futures flow, big trades, absorption/refill research, cross-venue confirmation and a Zone Reaction Engine with outcome tracking. The goal is to turn complex market behavior into structured, measurable evidence. Full case study on GitHub.

LLM, News Intelligence System

Controlled LLM intelligence layer for financial and crypto market analysis. Built around structured evidence, bounded context, scope controls, source-grounded News Intelligence, risk flags, contradiction detection, schema-validated reports and post-report outcome review. The LLM acts as an analytical component — not an autonomous trading agent. Full architecture and workflow on GitHub.

Production ML Engineering, Monitoring

Production engineering layer for moving time-series ML from research into a controlled runtime. Includes immutable model packages, frozen feature schemas, deterministic parity checks, SHA256 integrity, six-model inference, atomic SQLite persistence, worker lifecycle, outcomes and forward monitoring. Historical feature parity reached 1,461/1,461 eligible dates with zero feature-value mismatch. Full production case study on GitHub.

DanilaKuznetsov