Most AI projects die as demos. Let's scope one that doesn't.
Describe a process that eats your team's week. The scoping agent below reads it the way I would on a first call — signals, architecture, the thing most likely to kill it — and gives you an honest answer, including "this doesn't need AI."
The gap is never the model
You have probably run an AI experiment already. Something that impressed people in a meeting, then never touched real data.
What breaks a project is everything around the model: integration with the systems you actually depend on, the inputs nobody anticipated, keeping a human in the loop where the stakes demand one, and knowing what it costs per run at real volume. Closing that gap is the entire job — and it is a delivery problem before it is a machine-learning one.
The agent above is a small demonstration of the same instinct: it does the cheap pass first, commits expensive effort only where the cheap pass found something, and tells you when the answer is no.
Three ways to work together
AI agents & automation
Custom agents for document processing, research, support and internal workflows — RAG architectures, LangChain, the Anthropic API — wired into your existing stack rather than bolted beside it.
AI systems consulting
An independent read on where AI actually pays off, a cost and ROI model with real numbers in it, and a delivery plan — before anyone writes code. Useful if you have been sold AI for its own sake before.
Meta-agent & orchestration
Agent systems that build, test and deploy other agents, with cost tracking and profitability modelling designed in from the first commit rather than discovered in the invoice.
How the work runs
Discover
A working session on your real process — not a discovery questionnaire. Output: a map of where AI removes friction and where it would just add a moving part.
Prototype
A working version in weeks, tested against your real data and its real edge cases. If it fails here, you lost weeks instead of quarters.
Deploy
Production hardening, monitoring, integration and handover — including cost per run, so the finance conversation happens before rollout rather than after.
Iterate
Systems drift as your data and users change. Ongoing refinement, or a clean handover to your team — whichever you actually want.
Built, not slidewared
Finding the twelve good seconds in sixty gigabytes of footage
An end-to-end AI video pipeline that scans hours of raw footage and surfaces the moments worth keeping. It runs a fast, cheap vision pass to find candidates, then an expensive precision pass only where the first pass hit — the architectural decision that makes the economics work at all. The same two-pass instinct drives the agent at the top of this page.
Built solo end to end: architecture, backend, review interface, and a cost model that recalibrates itself against real runs instead of guessing. Human-in-the-loop review throughout, because full automation would have been the wrong answer here.
Who you would be working with
UsturAI is the AI practice of a Tallinn engineer with two decades of running IT for organisations that cannot afford downtime — including an ERP implementation in recent years — now working full time in applied AI engineering.
That combination is the point. Building an agent is increasingly common. Knowing how a real organisation absorbs a new system — who resists it, what breaks in month three, which integration nobody mentioned at kickoff — is what decides whether it ships. You get the person who does the work, not an account manager relaying to a delivery team.
Working stack: Node.js, Python, RAG architectures, LangChain, FAISS, the Anthropic API, PyTorch — and the judgement to say when none of them is the answer.
What this page does with what you type
Selling GDPR work while running a careless lead form would be a poor advertisement, so here is the whole of it.
The assessment, by default
The scoping agent runs as JavaScript inside your browser. Your text is never transmitted, never logged, and never reaches a server — mine or anyone else's. Close the tab and it is gone.
If you tick "full model"
Your description is sent to Anthropic's API in the United States for that single request, and the answer comes straight back to you. I do not store it. The transfer relies on Anthropic's Standard Contractual Clauses. If that is not acceptable for your material, leave the box unticked — the local assessment stays available and is no less honest.
If you ask for a reply by email
- Stored: your email address, the description you wrote, and the assessment you were shown.
- Why: to answer you. Article 6(1)(a) — your consent, given by ticking the box.
- Where: Cloudflare storage in the EU, plus one draft-generation call to Anthropic (US, SCCs).
- How long: 180 days, then automatic deletion. This is enforced by the store itself, not by me remembering.
- Never: sold, shared, added to a newsletter, or fed into an automated sequence. A reply is written by a person.
Your rights
Access, correction, deletion, portability, and withdrawal of consent at any time — write to hello [ät] usturai.com and it is done the same day. You may also complain to the Estonian Data Protection Inspectorate (Andmekaitse Inspektsioon).
Cookies
None. No analytics, no tracking pixels, no consent banner — because there is nothing to consent to.
Ran the assessment? Let's pressure-test it.
No deck, no discovery fee, no pitch. Thirty minutes on whether the thing is worth building — including an honest no.
Start a conversationor write directly to hello [ät] usturai.com
Sama perearstikeskuse IT-partner alates 2005. aastast.
Kaks aastakümmet ilma dramaatikata. Nüüd koos praktilise GDPR-inventuuriga, mis lihtsustab teie andmekaitsespetsialisti tööd — mitte ei lisa kaustatäit paberit.
Kaks asja peavad korraga töötama
Esiteks peab igapäevane IT lihtsalt toimima. Arvutid, tarkvara, varundus, turvalisus — ilma et keegi keskusest peaks nende pärast oma päeva ümber korraldama.
Teiseks karmistuvad andmekaitsenõuded. Ja siin on nüanss, mida sageli tähele ei panda: suuremas keskuses ei tohi andmekaitsespetsialist ja IT-vastutaja olla sama inimene — see on huvide konflikt. Te ei vaja IT-partnerit, kes võtaks endale ka DPO rolli, vaid partnerit, kes teeb teie DPO töö oluliselt lihtsamaks.
Nii me ka positsioneerime end: IT-partner, mitte teie andmekaitsespetsialist.
Mida me teeme
IT-haldus ja tugi
Töökohtade ja serverite hooldus, tarkvarauuendused, varundus ja taastamine, võrgu- ja turvaseadistus. Kohandatud tervishoiuasutuse eripärale — sh sellele, et vastuvõtt ei saa seisma jääda.
GDPR-inventuur
Ülevaade sellest, kus patsiendiandmed tegelikult liiguvad ja kes neile ligi pääseb. Tulemus on dokument, mida saab kasutada, mitte üldsõnaline vastavusraport.
Lepingud ja mõjuhinnangud
Artikkel 28 volitatud töötleja lepingud, DPIA ettevalmistus ja tehniliste turvameetmete kirjeldused — koostatud nii, et teie DPO saab neile toetuda, mitte neid uuesti kirjutada.
Miks meie
Sama perearstikeskus on klient alates 2005. aastast — katkematult. Kakskümmend üks aastat tähendab, et oleme näinud, mis läheb katki teisel, viiendal ja viieteistkümnendal aastal. Ja et suhe on need kõik üle elanud.
Kaks aastakümmet IT-d organisatsioonides, kus seisak maksab. See kogemus loeb just siis, kui midagi on vaja välja vahetada või üle viia ilma tööd seisma panemata — mitte siis, kui kõik niikuinii töötab.
Väike partner, otsekontakt. Te ei helista kõnekeskusesse ega selgita oma olukorda iga kord uuele inimesele. Sama inimene, kes teie süsteemi tunneb, võtab ka telefoni.
Räägime, kus teie keskuse IT ja andmekaitse täna seisavad
Esimene vestlus on tasuta ja mittesiduv. Kui selgub, et teil on kõik korras, ütleme ka seda.
Võta ühendustvõi kirjuta otse tere [ät] usturai.ee