lalayants RU

Kirill Lalayants

Data Scientist · ML & Finance

NES · Skoltech · ITMO

I’m a Data Scientist at Sberbank CIB, applying ML to trading and pricing, with a strong interest in finance and economics. At Yandex Lavka, I optimized discounts (+5% Front Margin) and automated experiment setup. Previously, I conducted research at ITMO and led an R&D team.

CV (PDF)GitHubLinkedInEmail

Experience

  • Sberbank CIB

    May 2026 — present

    Data Scientist

    • Develop and implement RL agents for pricing and trading desk optimization.

  • Yandex Lavka

    Jan 2026 — May 2026

    Data Analyst

    • Optimized discount levels using price elasticity estimates, increasing Front Margin by 5%, as confirmed by a large-scale A/B test.
    • Automated experiment setup and approval, reducing preparation time from several hours to a few minutes.
    • Built infrastructure for collecting and analyzing experiment results, along with dashboards for price elasticity analysis.
    • Automated product selection for experiments based on sales volume and popularity.

  • ITMO University

    Jul 2022 — Jun 2025

    Research Engineer

    • Built a SLAM system from scratch for GPS-denied indoor UAV flights: adapted the algorithm and achieved 25 FPS on Jetson Xavier in real-world flights.
    • Implemented RL for a manipulator in simulation, achieving a 95% grasp success rate and reducing collisions.
    • Developed tools to calibrate tool and camera poses on a real manipulator, with errors below 1 mm.
    • Developed socket segmentation from RGB-D point clouds, plane estimation and distance calculations in the robot coordinate frame; the system was used for automated connector insertion on a real manipulator.

  • BRIZ Project

    Dec 2023 — Jul 2024

    Team Lead

    • Led the development of an autonomous docking system for two UAVs: achieved a docking success rate above 90% on real hardware and delivered a working prototype to the client.
    • Assembled the team, set the roadmap, approved technical approaches, ran tests and worked with company leadership.

Skills

ML & Deep Learning
PyTorch, neural networks (MLP), generative models (VAE/CVAE), RL, gradient boosting (CatBoost), scikit-learn, Optuna, hyperparameter tuning and model validation
Econometrics & experiments
A/B testing, DiD, uplift modeling, price elasticity estimation, time series econometrics, panel data
Quantitative finance
Portfolio optimization, asset pricing, CAPM and factor models, risk modeling, VaR/CVaR, duration and fixed income analysis
Data & analytics
Python, SQL, Pandas, NumPy, SciPy, Jupyter, Spark, Hadoop, Presto
Engineering tools
Git, Docker, configuration and experiment automation, event logging and metric pipelines, analytical dashboards
Robotics
Computer vision, image and point-cloud segmentation, RGB-D, 6D pose estimation, SLAM, ROS2, tool and camera calibration, MuJoCo, Gazebo, PyBullet, NVIDIA Jetson Xavier
Project leadership
Team formation, roadmapping, technical decision review, test coordination, client communication

Education

  • New Economic School

    2025 — 2027

    Master’s program · AI & FinTech

  • Skoltech

    2025 — 2027

    Master’s program · Data Science

  • ITMO University

    2021 — 2025

    B.Sc. · Robotics & Artificial Intelligence

Patent & software registration

Contributions & achievements

  • Ultralytics contributorFixed a bug that occurred when moving models to a TPU.
  • StarLine Hackathon winner“Autonomous Future”, 2023.
  • English · AdvancedTOEFL iBT · 106/120.

Contact

lalayantskirill@gmail.com@kirill_lalayants