Ship faster.
Build smarter.

AI & Automation consultant helping startups and enterprises build intelligent workflows, integrate LLMs, and deliver products faster.

10+ years building products

AI IntegrationProduct StrategyTech Advisory

How I Can Help

I combine deep technical expertise with product thinking to help you leverage AI and modern workflows.

AI & Automation

Integrate LLMs and automate workflows to save time and reduce manual work. From chatbots to intelligent document processing.

  • LLM Integration (Claude, GPT)
  • Workflow Automation (n8n, Zapier)
  • Custom AI Solutions
  • Process Optimization

Product Management & AI Adoption

Strategic product guidance with a focus on AI adoption. Help your team ship faster with modern practices.

  • Fractional PM/CPO
  • AI Adoption Strategy
  • Product Roadmaps
  • Team Coaching

IT & Tech Consulting

Technical guidance for architecture decisions, code quality, and technology selection. Build on solid foundations.

  • Tech Stack Advisory
  • Architecture Reviews
  • Code Quality Assurance
  • Development Oversight

Not sure what you need? Let's talk about your challenges and find the right solution together.

Book a Free Call

Live reference

An AI agent fleet that works while you sleep

I don't just advise on AI — I run a production multi-agent Claude fleet every day. A 24/7 orchestrator coordinates parallel developer agents that code, review, research, and report autonomously. Everything below is running right now, not a concept.

Orchestrator24/7codes & opens PRsreviews & testsresearches & reportswatches CI & deploysOne coordinator, many hands — tasks flow out in parallel, results flow back verified.

24/7 Orchestrator

A coordinator agent assigns work, deconflicts ownership, monitors health, and escalates decisions to a human phone via Telegram — around the clock.

Parallel Developer Agents

Multiple agents work simultaneously on separate tickets and branches: implementing fixes, opening pull requests, and watching CI through to merge and deploy.

Deep Tool Integration

Agents operate the same tools a team does: Linear tickets, GitHub PRs and reviews, Microsoft Teams messages, Telegram notifications, and voice summaries.

Overnight Autonomous Work

Hand over an acceptance checklist in the evening; the fleet implements, self-tests against live systems, and documents evidence — done by morning.

Research & Reporting

Deep-dive market or technical research with source-graded findings, delivered as a structured report — for example a full Hungarian AI-consulting price analysis, produced overnight.

Self-Verifying Quality

Adversarial review agents verify each other's findings, watchers auto-merge only on human approval, and every claim gets checked against the system of record.

A real night shift

Recently the fleet was handed a 9-point acceptance list at midnight — verify a live voice-call stack, fix layout determinism, prove each item with its own tests. By 5 AM, agents had verified the items against a live bridge, fixed the bugs they found, merged the fixes, and left a written evidence trail. The same night, another agent produced a complete market research report on Hungarian AI consulting rates. This is what AI-native operations look like.

Built work

Jarvis — call your AI from your wrist

A personal AI assistant you can literally phone. An iPhone and Apple Watch app pair, connected to the agent fleet on a Mac: Claude rings you, you talk, it acts.

YouiPhone ringsMac bridgeAI fleetYou speak — the call travels through the phone and the Mac to the fleet, and the answer comes back as voice.

Live calls with Claude

Real CallKit calls — the phone rings like a normal incoming call, and you have a live spoken conversation with the AI.

Apple Watch companion

Answer, mute from the wrist, and glance at fleet status via watch complications — without touching the phone.

Full voice pipeline

Speech-to-text and text-to-speech chained end to end, so conversations flow naturally in both directions.

Mac bridge

A REST API on the Mac exposes the whole agent fleet — session status, context levels, live summaries — with self-healing tunnel connectivity from anywhere.

Push notifications

APNs push keeps the phone and watch informed the moment an agent finishes work or needs a decision.

Decision queue

Everything waiting on a human — approvals, choices, blockers — surfaces in one ranked queue, answerable from the phone.

Built in Swift (SwiftUI, CallKit, WidgetKit) with a Python bridge — designed, implemented, and maintained together with the AI fleet itself.

Pricing

AI consulting, transparent rate

Most consultancies hide their rate behind a sales call. Mine is public.

AI Consulting Hour

69 420HUF / hour + VAT

Kiss Gergő EV · invoice provided · remote or Budapest on-site

  • 1:1 consultation on your actual processes — where AI genuinely helps, and where it doesn't
  • Live demo of a production multi-agent fleet, not slides
  • Concrete adoption recommendations: tools, integration path, first wins
  • Written follow-up summary with action points after every session
Book a Consultation

Free 30-minute discovery call if you share your thoughts and context in advance with your inquiry.