Олексій

OLEKSII YEROMKIN
D A T A S C I E N T I S T /A N A L Y S T /M A C HI N E L E A R N I N G E N G I N E E R

Data Scientist and Engineer with deep expertise in financial data analytics and time‑series
modeling. I design and implement end‑to‑end data solutions—spanning acquisition,
preprocessing, advanced modeling, and results interpretation—to uncover actionable insights,
resolve complex business challenges, and automate mission‑critical workflows at scale.
My technical approach combines rigorous engineering discipline with a scientific research
mindset. I actively explore and apply a versatile toolkit that includes artificial intelligence,
classical machine learning, and mathematical modeling, always striving to build robust,
explainable, and production‑ready systems.
Driven by intellectual curiosity and a passion for innovation, I seek to bridge the gap between
cutting‑edge research and tangible business value—delivering solutions that are not only
performant but also deeply grounded in sound methodological principles.

TECH Python, numpy, pandas, scikit-learn, statsmodels,
LangChain, PyTorch, MongoDB Atlas, MS SQL, Postgres
SQL, DBeaver, matplotlib, streamlit, plotly, bs4, FastAPI,
GitHub/GitLab, Hugging Face, Jupyter Notebook, AWS,
Spark, Coolify, Docker, GitHub Copilot etc.

EXPERIENCE DATA SCIENTIST / QUANT ENGINEER
NDA 09.2025-06.2026
Served as the Lead research-engineer responsible for
architecting and deploying quantitative trading strategies
(including SMA, Martingale, etc). Oversaw rigorous
backtesting frameworks and hyperparameter optimization,
leveraging advanced techniques such as Monte Carlo
simulations, Bollinger Bands, RSI, Random Search, and
Bayesian Optimization. Executed deep-dive analyses of
cryptocurrency markets utilizing econometric tools
(regression analysis, PACF, and GARCH/ARCH models).
Successfully transitioned these strategies into fully
operational trading environments designed for multi-investor
portfolios, encompassing end-to-end API integrations,
dynamic UI dashboards, and scalable database architectures.

AI/ML ENGINEER
NDA 05.2025-08.2025
Designed and deployed a Retrieval-Augmented Generation
(RAG) system to enable semantic search and Q&A over
proprietary company knowledge. Utilized Text Generation
Inference (TGI) to ensure low-latency and highly efficient LLM
inference. Significantly accelerated the onboarding process
for new developers and reduced the time spent searching for
internal documentation and code standards.

DATA SCIENTIST / ML ENGINEER
Hstepanalytics 05.2024-04.2025

Spearheaded the development of an MVP that seamlessly
fused predictive analytics with LLM-powered robo-consulting
for automated investment solutions. This platform delivered
comprehensive financial insights by integrating
conversational AI (LLMs) for user interaction with robust
forecasting engines—including Linear Regression (LR), LSTM,
and Prophet. Additionally, led exploratory R&D initiatives to
refine existing analytical methodologies; conducted
experiments to uncover novel market patterns, optimized
shortest-path algorithms, performed in-depth market
microstructure analysis, and enhanced adaptive gradient
techniques to boost model efficacy.

DATA SCIENTIST / DEVELOPER
Freelance 01.2023-05.2024
Architected and maintained robust ETL pipelines for
extracting, cleaning, and aggregating raw data from diverse
sources. Developed and deployed machine learning models
(XGBoost, Linear Regression, Scikit-Learn algorithms) to solve
actionable business problems (forecasting, classification,
optimization). Conducted comprehensive Exploratory Data
Analysis (EDA) and feature engineering, ensuring models
were interpretable and seamlessly integrated into clients
workflows.

DATA SCIENTIST / BI DEVELOPER
Helix Global LLC 07.2021-12.2022
Engineered high-performance ETL pipelines to streamline
analytical computations, drastically accelerating the
throughput of the business intelligence department.
Contributed to the design and deployment of a client-centric
recommendation engine that utilized autoencoders,
customer-product affinity matching, and dimensionality
reduction techniques. This initiative significantly increased
client on-site engagement and session duration, directly
enhancing user retention and platform stickiness.

PORTFOLIO Linear regression for predictions simple trend

GARCH for predict market volatility

Search path in 3d

EMAIL [відкрити контакти](див. вище в блоці «контактна інформація»)
WA / TG [відкрити контакти](див. вище в блоці «контактна інформація»)
LN Oleksii Yeromkin