OpenAI
2026 –- Prism: Engineering lead across nearly all of Prism (prism.openai.com), owning product and platform architecture end to end.
- Codex: Building OpenAI's agentic coding platform across product and core infrastructure.
Aleksey Smolenchuk
Software engineer · San Francisco
17 years of experience building web applications, applied AI, developer tools, and infrastructure at early-stage startups and global technology companies.
Experience
Download resumeCrixet
2025 –- Founding engineer of a real-time collaborative LaTeX editor; acquired by OpenAI (2025).
- Led a rearchitecture of the backend, CRDT-based sync engine, and sandboxed compilation infrastructure, cutting infrastructure costs by 92% through compute and pipeline optimization.
- Owned product features and infrastructure end to end, sustaining rapid user growth through launch and acquisition.
Netflix
2023 – 2025- Built and operated world-scale Node.js services serving critical data to millions under strict latency and uptime SLAs.
- Architected developer-platform components and CI/CD that measurably improved build and deploy velocity across engineering.
- Drove software supply-chain security across the JavaScript platform.
Opendoor
2019 – 2023- Owned and scaled a mission-critical cross-org Node.js GraphQL gateway serving critical production traffic at 99.99% reliability.
- Established engineering-platform standards, automation pipelines, and shared tooling adopted org-wide.
Uber
2014 – 2019- Technical leader for the internal developer platform: frontend frameworks (JavaScript, Node.js) powering hundreds of web apps company-wide.
- Launched both major customer-facing product features as well as developer experience + internal tools and platforms at scale.
- Engineering Manager (2017 – 2018): Led a team delivering developer-productivity processes and systems.
Branding Brand
2012 – 2014- Architect lead on mobile e-commerce platforms for world-leading retailers.
Ansaldo STS (Hitachi Group)
2010 – 2012- Built Linux systems- and kernel-level, safety-critical embedded software for train-control systems.
Education
University of Pittsburgh2007–2010
B.S. in Computer Science
Skills & languages
Languages: JavaScript/TypeScript, Python, Previously used: Go, Rust, Lua, C, C++
AI & ML: LLMs, fine-tuning / post-training (HF stack, LoRA/PEFT), agentic systems, RAG, evaluation, prompt engineering
Cloud & Infrastructure: AWS, Azure, IaaS, Kubernetes, Docker, Terraform, Cloud networking, supply-chain security, GraphQL, gRPC, monorepos, CI/CD pipelines, observability, operation & scaling
Frontend: React, Next.js, Node.js, a11y, i18n, browser APIs, web performance, web security
Product & Systems: Product engineering & design, developer platforms, rapid prototyping; Linux (kernel/systems-level), embedded & safety-critical software; high-QPS microservices, distributed systems
Spoken languages: English (native), Russian (native), Belarusian
A few personal projects
All on GitHubsimple-llm-finetuner
A simple interface for fine-tuning language models.
llavavision
A “Be My Eyes” web app using a local vision-language model.
wasm-gpt
A browser-based chatbot built with ggml and WebAssembly. Runs locally, without API keys or a backend.
chrome-mcp
Let LLM agents operate Chrome without taking screenshots.
yapyap
Fast, simple push-to-talk dictation.