Pavel Rybakov

Founding-engineer style builder with 20+ years across engineering, delivery, infrastructure, and organizational transformation. Building software systems and engineering workflows that raise the level of abstraction.

🌐 English C1-C2 📍 Open to London on-site 💰 ~$8,000 + equity
What I Bring
Languages
Clojure, TypeScript, Node.js
Frontend
React
Infra & Delivery
Kubernetes, Terraform
System Design
Event-driven, config-as-code, codegen
AI-Assisted Dev
Claude Code, Codex, agent workflows
Leadership
Startup to 1,500-person orgs
Experience

Salmon.ph

Operations Engineering Manager
Feb 2023 — Present
  • Redesigning engineering workflows around agent-driven development and practical software delivery
  • Built reusable agent-ready templates with React + Clojure foundations
  • Built a Kubernetes deployment skill enabling required K8s and Terraform changes in minutes
  • Helped reduce delivery timelines from months to weeks
  • Within one month, 5 systems were already being developed on top of the templates

Guides in the City

Co-founder / CTO
Feb 2016 — Present
  • Built a semi-automated excursion sales business across 100+ cities in Russia
  • Designed event-driven architecture with receptors, handlers, and executors in Node.js
  • Reduced excursion sales cycle from ~2 days to 4 hours
  • One manager could sell 300–400 excursions/month; network of 1,000+ guides

PJSC T Plus

General Director, Internal IT Company
2013 — 2016
  • Led IT transformation in a major energy company
  • Reorganized ~1,500 employees from branch-based silos into function-based specialization groups

System Operator of the Unified Power System

Head of Software & Hardware Systems
2011 — 2013
  • Reliable operation of software and hardware systems across 50 branches

Enel Russia

Head of IT Operations
2008 — 2011
  • IT operations for central office and 4 power plants in a distributed environment
Why CodeSpeak

Spec-first, agent-driven

At Salmon, I designed templates and deployment skills used directly by agents. The hard part is encoding enough structure and intent for agents to reliably produce useful systems.

Low-code intuition

Code is often not the bottleneck. The real bottleneck is whether behavior, constraints, and intent are specified clearly enough for humans and tools to act correctly.

Building from ambiguity

I design the architecture, define APIs, and build the execution model. The hard part is always turning fuzzy requirements into a system with stable boundaries.

Operating across levels

I can discuss abstractions, architecture, delivery workflows, and organizational structure — then go and build the thing.

Education
Systems Engineer
Vladimir State University
2003
Economist
MGIMO University
2009
Positioning
I am not a conventional FAANG-style specialist optimizing one narrow layer of the stack. I am strongest when the problem is ambiguous, the system needs clearer abstractions, and the path to better software delivery runs through better specifications, better workflows, and practical engineering judgment.