Innovedi · AI agents and automation

AI where it pays off. Plain automation where it's enough.

We help companies find the business problems worth solving, then apply the right tool: AI agents where they make a real difference, standard software automation where that does the job better. We build with your team and hand over working systems you own.

Our approach

Start with the problem, not the technology

AI is powerful, and it is not a silver bullet. The useful question is not "where can we add AI?" but "what is slowing this business down, and what is the simplest thing that fixes it reliably?"

  • Problem first.

    We start from how your business actually runs: the processes, the systems, the friction.

  • The right tool for each job.

    AI agents where judgment, language or messy inputs are involved. Rules, integrations and conventional software where they are cheaper, faster and more predictable. Most real solutions combine both.

  • Built for production.

    Evaluation, guardrails and human review are designed in from the start, so what works in a demo keeps working with real users and real data.

What we do

From first map to working systems

Most engagements start with a short mapping phase and continue into building. Each phase stands on its own, with a clear scope and outcome.

/map

Opportunity mapping

We look at your processes, data and any AI work already underway, and rank where AI or automation will pay off. You get a prioritized plan and an architecture your team could build from.

/agents

AI agents and knowledge systems

Agents connected to the systems you already use, and shared company knowledge that people and agents can both search and update, in stores you control.

/automation

Software automation and integration

When the answer is not AI: workflow automation and system integration that remove manual work and connect the tools you already run.

/leadership

Fractional engineering leadership

Senior technical leadership without a full-time hire: architecture and build-versus-buy decisions, hiring the first engineers, and adopting AI coding agents with practices that hold up.

Process

Start small, build together, hand over

Every engagement has a beginning and an end. We work in bounded phases, and after each one you decide what comes next.

  1. 01

    Discover

    A short, clearly scoped start: how your business runs, which systems it depends on, where the friction is.

  2. 02

    Map and rank

    Together we map the problems worth solving, rank them by value and effort, and agree on phases with clear outcomes.

  3. 03

    Build together

    We build alongside your team, inside your real workflows, in short cycles you can see and test.

  4. 04

    Hand over

    Working software, documentation and processes your team can run. Every phase ends in a state someone else could pick up.

  5. 05

    Stay available

    Reviews, advice or a next phase when you want them. Whether you continue is your call.

What you keep

No lock-in

When the engagement ends, the capability stays with you. Keep working with us because it helps, not because you have to.

  • Your code.

    Everything we build lives in your repositories and runs on your accounts.

  • Your data.

    Your data and company knowledge stay in systems you control.

  • Your processes.

    Documented, and practiced by your own team before we step back.

  • Your options.

    Each phase stands on its own. Nothing ties you to us, or to a single AI model provider.

Open source

Built in the open

The clearest example of how we work is the software we publish: your knowledge in your own store, no vendor in the middle.

Cerefox

User-owned shared memory for AI agents: a knowledge base that Claude, ChatGPT, Cursor, Codex and your own agents can all read and write, in a database you control.

About

About Innovedi

Innovedi was founded by Fotis Stamatelopoulos. He has spent his career scaling software systems and the engineering teams behind them, first running his own consultancy in Europe, then leading engineering organizations of more than 100 engineers at a US public technology company. Today he builds AI-agent systems and the teams that ship them.

Contact

What would you like to take off your plate?

Tell us about the problem, the systems involved and what a good outcome looks like. The first conversation is about understanding your situation.

hello@innovedi.com