Open to opportunities

Vladimir Aleksenko

@vladimiraleksenko

Message

I build production machine learning systems and decision-ready models from complex, noisy data.

Armenia

Message

What I'm looking for

I'm looking for research-driven work with incomplete information, where I can determine what questions the data can answer, build production-ready ML systems, and translate modeling results into decisions.

I've shipped three ML models for Deep Origin's production simulation platform, owning problem framing, dataset construction, architecture design, evaluation, and integration. I also reduced inference cost by replacing per-query transformer inference with a precomputed embedding database and vector similarity search.

At Magical Labs, embedded with the research team delivering for Deep Origin, I built self-supervised embeddings for 3D structural data and an unsupervised pipeline for large unlabeled time-series data. I turn open-ended research questions into concrete modeling approaches and explicit client recommendations.

Earlier, at the Research Center for Molecular Mechanisms of Aging and Age-Related Diseases, I ran large-scale sequence similarity retrieval on HPC clusters and GCP, built data-quality workflows across incomplete reference sources, and applied probabilistic inference, clustering, and dimensionality reduction to biological datasets.

Across data science and ML engineering, I work end to end: deciding what can be learned from the data, designing robust evaluations when standard metrics mislead, and delivering systems that are practical to operate at scale. I mentor junior engineers, lead senior-staffed projects, and communicate results as decisions rather than metrics.

Experience

Work history, roles, and key accomplishments

Senior ML Engineer / Data Scientist

Deep Origin

Jan 2024 - May 2026 (2 years 4 months)

Owned the full modeling lifecycle for ML components of a production simulation platform, shipping three models to production. Designed compact neural architectures, reframed ill-conditioned regression as classification, and cut inference costs with embedding databases.

PyTorch TensorFlow CNN Focal Loss Bayesian Optimization Python Docker CI CD MLFlow GCP Streamlit

ML Research Engineer / Data Scientist

Magical Labs

Jan 2022 - Dec 2024 (2 years 11 months)

Drove research for modeling approaches later taken to production, building self-supervised embeddings via contrastive regression and designing unsupervised pipelines for time-series structure recovery. Delivered research results directly to clients.

Self Supervised Learning Contrastive Learning PCA Clustering Time Series Analysis Python PyTorch

Research Scientist, Computational Biology

Research Center for Molecular Mechanisms of Aging and Age-Re

Jan 2018 - Dec 2022 (4 years 11 months)

Conducted large-scale similarity search and retrieval on HPC clusters and GCP, designed multi-stage filtering funnels, and performed maximum-likelihood inference for phylogenetic reconstruction. Merged heterogeneous datasets and propagated labels to improve annotation quality.

Developer Intern

Yandex

Aug 2021 - Oct 2021 (2 months)

Worked on training-data quality for a speech recognition system, detecting noisy and mislabeled examples and developing automated classification to correct them.

Data Quality Classification Python

ETL Developer

GlowByte Consulting

Jan 2017 - Dec 2018 (1 year 11 months)

Migrated a production credit-scoring data pipeline from Oracle to Hadoop, translating SQL dialects and building a Python orchestrator for parallel computation that reprocessed two years of data in two months.

SQL Oracle Hadoop Python ETL Data Pipelines Parallel Computing