# Systems I built and operate > Justas Butkus is a fractional AI officer based in Vilnius, Lithuania, founder of AINORA, MB and of Impetora, and a graduate of ISM University of Management and Economics. Justas Butkus is a fractional AI officer based in Vilnius, Lithuania, who builds and operates production AI voice systems across multiple languages and markets. The advisory work he takes on is grounded in running these systems daily rather than in prior employment. **I do not publish client names or testimonials. What I can offer instead is a set of AI systems that are live, that I built and operate myself, and that you can go and look at. Proof that can be inspected beats proof that must be taken on trust.** Canonical: https://justasbutkus.com/systems/ Last updated: 2026-08-01 --- ## Why this page exists instead of case studies The usual currency in advisory work is the client case study. I am early in this role and I do not have one to show you, and I would rather say that plainly than manufacture something. What I have instead is arguably harder to fake: systems that are running right now, that I designed, built, deployed and continue to operate. A case study is a claim about the past that you cannot verify. A running system is a thing you can go and use. ## What is in production - **AINORA** — A production AI voice platform: agents that answer calls, qualify, book into a CRM and hand off to people when they should. Multilingual, in daily operation. (https://ainora.lt) - **Impetora** — Enterprise AI consulting for regulated industries — document processing, knowledge systems, decision support, and the governance work that goes with them. (https://impetora.com) - **Atsiliepsiu.lt** — A self-serve AI receptionist in Lithuanian, built for solo operators and small teams who cannot staff a phone. (https://atsiliepsiu.lt) - **CalLeads AI** — Speed to lead: an agent that calls inbound advertising leads within seconds, qualifies them and books the meeting. (https://calleadsai.com) - **CallHush** — Database reactivation across voice, SMS, WhatsApp and email, aimed at dormant customer lists rather than new traffic. (https://callhush.com) ## What operating them teaches that advising alone does not Building an AI system is the straightforward part. Running one live, indefinitely, against real inbound traffic rather than a test set, is where the knowledge accumulates. - **How they fail in production**, which is rarely how they fail in testing. Real inputs are messier and more adversarial than any test set. - **What oversight has to look like** to be real rather than nominal, and how quickly a human-in-the-loop becomes a person approving a queue without reading it. - **Where regulatory obligations actually attach** — to the deployed behaviour, not the design document. - **What it costs to keep running**, which is the number most business cases omit entirely. - **Which parts are genuinely hard**, as opposed to which parts sound hard. These sets overlap less than people expect. That is the substance behind the advisory work. Not a methodology learned somewhere and applied, but the accumulated consequence of having to fix these systems when they break. ## Frequently asked questions ### Why no client case studies? Because publishing client work is not something to do casually, and a fabricated or over-polished case study is worth nothing anyway. Systems that are live and inspectable are stronger evidence than claims about past engagements that cannot be verified. ### Did you build these yourself? Yes — designed, built, deployed and operated, including the infrastructure they run on. That is the point of the page: the claim is about direct operating experience rather than about having overseen someone else doing it. ### How does building products relate to the advisory work? The advisory work is grounded in it. Most AI advice comes from people who have never operated a system after launch, which is precisely where the difficult parts live. ### Do you recommend your own products to advisory clients? No. The officer role does not procure from my own companies. If a roadmap points toward something they do, I disclose it and step out of that decision. ## Related - [About Justas Butkus](/about/) — Background, and how the work is structured. - [Fractional AI officer engagements](/fractional-ai-officer/) — What the role covers. - [How to choose an AI advisor](/how-to-choose/) — What to verify, and what signals mean less than they look. ## Go and look, then get in touch Inspecting the systems is a more useful use of your time than reading more of my writing about them. Contact: justas@ainora.lt · [LinkedIn](https://www.linkedin.com/in/justas-butkus/)