yvan_lokhmotov --init Cambridge, UK · A-Levels · Software Engineer

YVAN
LOKHMOTOV

A-level student at Hills Road Sixth Form, Cambridge — shipping production ML systems and engineering hardware from scratch. Three years as a software intern at KRAI, with contributions to MLPerf Inference v3.0 and v3.1.

9×
Grade 9 GCSEs
3+
Years at KRAI
2
Physics Awards
VIEW_PROJECTS GET_IN_TOUCH
01 / about

Background

I'm reading Maths, Further Maths, Physics, Chemistry, and EPQ at Hills Road Sixth Form College, Cambridge. My work sits at the intersection of machine learning research, software engineering, and mechanical design — I'm drawn to problems that reward depth across more than one discipline.

At KRAI I've spent three years building benchmarking infrastructure used in industry-standard MLPerf Inference submissions. Outside of work I build for its own sake: a rigorous ML pipeline for F1 race prediction, and a direct-drive sim wheelbase engineered from first principles.

25–27
Hills Road Sixth Form College
Maths · Further Maths · Physics · Chemistry · EPQ
23–25
The Perse School, Cambridge
Nine grade 9s, two grade 8s · AS Computer Science — grade A (early)
skills.sh
$ls ml/
PyTorchTabNetpandasNumPyscikit-learn
$ls software/
PythonRustFull-stack web
$ls hardware/
CAD/3D PrintElectronicsFirmwareEncoders
$ls mlops/
MLPerfBenchmarkingCVNLP
$
02 / projects

Selected Work

proj_001 FEATURED
EPQ F1 Predictive Model
End-to-end ML pipeline predicting Formula 1 race finishing positions using neural networks and TabNet. Strict leakage prevention, feature engineering, and season-by-season backtesting.
PyTorchTabNetpandasscikit-learn
proj_002 FEATURED
DIY Direct Drive Wheelbase
Sim racing wheelbase engineered from scratch — repurposed hoverboard motor, custom CAD enclosure, PETG gyroid-infill structure, off-axis encoder, open-source FFBeast firmware. ~15 Nm torque.
CADElectronics3D PrintFirmware
proj_003
Advent of Code
Clean Python solutions to AoC challenges — emphasis on algorithmic correctness, efficiency, and readable implementation over brute-force approaches.
PythonAlgorithms
03 / work

Experience

Jul 2022 – Present
KRAI
Part-time · Cambridge, UK
Software Engineer Intern
  • Built reproducible ML benchmarking workflows across Arm-based hardware (NVIDIA Xavier, Raspberry Pi 3/4, Odroid N2+, T-Firefly RK-3399), contributing to official MLPerf Inference submissions v3.0 and v3.1 — a globally recognised industry benchmark.
  • Designed and built a full-stack web service to automate workflow command generation, reducing human error and significantly improving submission reliability.
  • Developed interactive data visualisations of benchmarking results for use in sales and marketing.
Feb 2026 – Present
Cambridge Private Tutors
Freelance · Cambridge, UK
Private Tutor
One-to-one Maths and English tuition for 10+ entrance and GCSE students. Designing structured, personalised lessons with regular assessment to build exam technique and subject confidence.
Oct – Dec 2024
Aspire Badminton
Volunteer · Cambridge, UK
Volunteer Coach
Coached children across a range of ages and abilities. Built patience, clear communication, and the skill of breaking down technique into accessible steps.
04 / recognition

Awards

GOLD
BPhO Senior Physics Challenge
Feb 2026 · Hills Road Sixth Form College

Gold award in the national Senior Physics Challenge — recognising outstanding performance in advanced problem-solving beyond the A-level syllabus.

BRONZE
British Physics Olympiad Round 1
Jan 2026 · University of Oxford / BPhO

Bronze as a Year 12 student competing predominantly against Year 13 entrants. Two-hour written exam on advanced problem-solving and mathematical reasoning beyond A-level.

ACADEMIC
Nine Grade 9s at GCSE
2025 · The Perse School, Cambridge

Nine 9s and two 8s across 11 GCSE subjects. AS Computer Science completed a year early — grade A.

INDUSTRY
MLPerf Inference v3.0 / v3.1 Submissions
2022 – Present · KRAI

Named contributor to official MLPerf Inference benchmark submissions from age 15, working as a part-time software intern on production benchmarking infrastructure.

05 / contact

Get in touch

Let's
build
something.

Open to internship opportunities, university conversations, and anything interesting at the intersection of ML and engineering.