CV
Lead Software Engineer who scales systems and engineering organisations. Currently owns the inference serving stack for generative AI on custom logarithmic-compute silicon — from tokenisation and speculative decoding through production serving in Rust.
Simultaneously drives developer experience improvements adopted across a 70+ engineer organisation, including strict type checking rollouts, monorepo tooling, and CI pipelines that cut infrastructure costs by over 30%.
8 years of experience shipping ML systems end-to-end, from training pipelines and multi-cloud infrastructure to low-level serving optimisation.
Skills
- Languages & Runtime: Rust (tokio, llama.cpp), Python (uv, pyright, ruff)
- ML & Orchestration: Flyte, Kubeflow
- Cloud & Infra: Google Cloud, AWS, Terraform
- Observability: Prometheus, Grafana, OpenTelemetry
- Web & Data: TypeScript, VueJS, MongoDB
Professional Experience
Lead Software Engineer — Tensordyne (Remote)
2025 – Present
- Core contributor to the inference serving engine for generative AI on custom ASIC
- Implemented speculative decoding, grammar-constrained sampling, and multimodal support (Rust, tokio, llama.cpp)
- Defined org-wide engineering standards (PR workflows, testing, dependency management)
- Drove monorepo consolidation and unified Python dependency management (uv workspace across 70+ engineers)
- Owned CI/build infrastructure for a polyglot monorepo (Rust, Python, C++)
- Introduced distributed build caching, stacked PR CI optimisation, and Dev Containers
- Reduced infrastructure costs by 30%+
Senior Software Engineer — Tensordyne (Remote)
2022 – 2025
- Led rollout of strict type checking and linting across Python codebase
- Organised engineering-wide best-practice sessions
- Built ML training pipeline orchestration and experiment tracking
(Flyte, Weights & Biases, Kubernetes on GCP)
2017 – 2022
- Built asynchronous job execution platform for ML inference workloads
- Enabled production-grade AI automation pipelines
- Developed hybrid & multi-cloud ML training strategy (GCP, Azure, AWS)
- Optimised spot instance usage with Kubeflow
- Built web-based annotation and management tooling
- Developed backend and IAM systems
- Coordinated nearshoring team
2015 – 2017
- CRM software engineering (Angular, PostgreSQL, RabbitMQ, PHP)
Education
2018 – 2020
Profile: Artificial Intelligence
Thesis: Improving Orientation Estimation in Sparse LiDAR Grid Maps
2017 – 2018
2012 – 2017
Thesis: Cooperative Collision Avoidance by Gaussian Processes
Outreach & Activities