Skip to content
View nefayran's full-sized avatar
🌍
Working from home
🌍
Working from home

Block or report nefayran

Block user

Prevent this user from interacting with your repositories and sending you notifications. Learn more about blocking users.

You must be logged in to block users.

Content in all repositories owned by your account will be closed.
Maximum 250 characters. Please don’t include any personal information such as legal names or email addresses. Markdown is supported. This note will only be visible to you.
Report abuse

Contact GitHub support about this user’s behavior. Learn more about reporting abuse.

Report abuse
nefayran/README.md

Rodion Kazennov

AI Solutions Architect · Forward Deployed Engineer

I build LLM agents into how engineering teams deliver software, and I measure whether they pay off.

  • Now: a headless agent pipeline that takes a ticket through spec, design, build, review and test, with people signing off at spec and QA. Connected to Jira, Confluence, Teams, Figma and cloud logs.
  • Before: tech lead for 12 engineers in Japan and Vietnam. Built an LLM pipeline that turns requirements into Cypress tests, and a Slack assistant that answers from the codebase with citations.

Open source

  • agentic-frontend-team: six Claude Code agents that take an RFP to a tested Next.js frontend.
  • code-graph-vs-grep: does a code graph save a coding agent tokens against Grep and Read? Measured on public repos with two Claude models, plus the MCP wrapper under test (write-up).
  • protocell-genesis: can a simulated primordial soup close itself into a vesicle? A WebGPU simulation with pass criteria fixed before the runs and a negative answer with a measured cause (live demo).

Merged upstream: avoid-ai-writing#352, vite-plugin-federation#261, code-coverage#600.

Claude Certified Architect – Foundations · allkeep.org · nefayran@gmail.com

Pinned Loading

  1. protocell-genesis protocell-genesis Public

    A WebGPU primordial-soup simulation with pass criteria fixed before the runs: the soup assembles, the vesicle never closes, and the cause is measured.

    TypeScript

  2. code-graph-vs-grep code-graph-vs-grep Public

    Does a code graph save a coding agent tokens compared with Grep and Read? A public benchmark, and cbm-lean, the MCP wrapper it tests.

    JavaScript

  3. agentic-frontend-team agentic-frontend-team Public template

    Six Claude Code agents that take an RFP to a tested Next.js frontend: analyst, architect, designer, engineer, QA, reviewer. Four skills, 15 practice RFPs.