Engagement models
Three ways to work with us.Pick your weapon.
Productized launches, bespoke builds, or custom-trained models — every engagement runs on the same rules: fixed scope, fixed price, senior engineering end-to-end, and a working demo in week one.
The tiers
Launch, Custom, or Fine-Tuning
Most clients start with Launch or Custom. Fine-Tuning is for the point where off-the-shelf models stop being good enough — we’ll tell you honestly whether you’re there.
Launch
Productized AI, fast
Pick a product, we configure it for your business. Live in 1–4 weeks, fixed pricing, monthly retainer covers hosting, model costs, and tweaks.
- Any product from the catalogue
- Configured to your brand & data
- Live in 1–4 weeks
- Monthly retainer from €50
Custom
Bespoke AI projects
RAG systems over your proprietary data, AI features embedded in your existing apps, and multi-step agents built for your exact workflow.
- Discovery sprint from €500
- Projects €5,000–€25,000
- 3–8 week delivery
- Ongoing improvement retainer
Fine-Tuning
Models trained on your domain
When off-the-shelf models plateau, we train custom models on your data — classification, extraction, and domain-specific generation.
- From €8,000, project-scoped
- Custom evals & benchmarks
- Deployed on your infra or ours
- For when APIs aren’t good enough
Decision guide
Which tier fits you?
A rough map. If you’re genuinely not sure, book the free call — we’ll place you in about ten minutes, and we won’t upsell you into a tier you don’t need.
Choose Launch if…
You have a well-known problem and want a proven system running this month, not a research project.
- You’re losing calls, leads, or hours to a problem one of our products already solves
- Your budget is under €5,000 and you want a predictable monthly cost
- This is your first AI project and you want a low-risk starting point
- You need it live in 1–4 weeks
Choose Custom if…
Your workflow doesn’t fit a template, or the AI needs to live inside software you already run.
- The answers must come from your proprietary documents or database
- You want AI features embedded in an existing app, portal, or CRM
- The process spans multiple steps, systems, or approval chains
- You’d rather de-risk with a €500 discovery sprint before committing
Choose Fine-Tuning if…
You’ve pushed prompting and RAG as far as they go, and the remaining errors are costing you real money.
- Off-the-shelf models keep failing on your domain’s jargon or formats
- You classify or extract from thousands of documents and accuracy is the bottleneck
- Compliance requires the model to run on your own infrastructure
- You have the data volume to make training worth it — and we’ll verify that first
The standard
What every engagement includes
Whichever tier you pick, the operating principles don’t change. This is the baseline — not the premium package.
- Fixed written scope and fixed price before work starts — no hourly billing, no scope drift, no surprise invoices
- A senior engineer end-to-end: the person on the discovery call is the person writing the code
- A working demo within the first week, so you’re evaluating software instead of slide decks
- Documentation and handover — you own the system, and any competent developer can pick it up after us
- EU/GDPR-conscious builds: data residency, retention, and access designed in from day one, not patched on later
- Honest advice, including “don’t use AI for this” — if a cron job or a form solves it, we’ll say so and save you the budget
Custom in practice
Custom project examples
Three representative scopes in the €5,000–€25,000 range, drawn from the kind of work we quote for legal, insurance, and hospitality clients. Your exact price comes from the discovery sprint.
RAG over case files
For a law firm
A retrieval system over 15 years of case files, pleadings, and precedents — scanned PDFs included. Lawyers ask in plain language, get cited passages back, and never see matters they don’t have access to.
- OCR ingestion of scanned archives
- Per-matter access control
- Cited answers in EN / EL
- On-premise or EU-hosted deployment
Indicative: €12,000–€18,000
AI inside your booking app
For a hospitality group
AI features embedded in an existing reservation platform: guest messages answered and triaged automatically, upsell suggestions at booking time, and reviews summarized for management weekly.
- Works inside your current codebase
- Multilingual guest replies (EN / EL / RU)
- Human handoff on low confidence
- No platform migration required
Indicative: €7,000–€12,000
Multi-step intake agent
For an insurance broker
An agent that takes a new claim or quote request from first contact to a complete file: collects details, extracts data from attached documents, checks for gaps, and hands the underwriter a ready dossier.
- Web form, email & document intake
- Automatic data extraction & validation
- Pushes to your existing broker system
- Full audit trail of every step
Indicative: €15,000–€22,000
Prices are indicative, not quotes. Every custom project starts with a discovery sprint (from €500) that ends in a fixed-scope, fixed-price proposal.
Not sure which tier? Ask.
Book a free 30-minute discovery call. We’ll place you in the right engagement model, map the first workflow worth automating, and give you a straight answer on cost — no obligation.
Want more depth first? Our paid discovery audit (€300–500) is credited in full against your first project.