// hi, I'm
Lead R&D Engineer at SoftServe. Founder of WhateverAI.
AI lowered the floor, our work now is the ceiling.
Anyone can get a first draft out of AI today. What separates a demo from something users can lean on is judgment: guardrails, evals, cost per inference, and knowing when not to use an agent. I have worked with AI in production since 2018.
AI Week 2026 talk: resources and 1:1 sessions →SRE triage agent. Turns an incident report into a severity-classified ticket with codebase analysis, runbook steps, and team notifications, with no human in the loop.
7-agent AI pipeline that writes, audits, and publishes SEO content. Each agent does one job, and anything that fails the audit goes back to the writer.
Every time an agent fools me, I write the rule down. The ones that can't fail no longer live in a prompt.
Sofía had a 26,000-line prompt and booked a visit for the wrong day. Intelligence wasn't what she lacked.
The auto-reply was perfect in testing. Before switching it on, I had it answer the real emails.
Get posts about AI, development, and the solo founder journey.