Machine learning & AI, delivered as a fixed-price project

Automate the decisions your team makes by hand. Running on your data in weeks.

Fraud you catch one case at a time. Demand you forecast in a spreadsheet. Documents someone reads and retypes. xval.ai builds the model that does it for every case, every day, and puts it into production in weeks. Fixed scope, fixed price, success criteria agreed before we start.

Book a free 30-min call Email hello@xval.ai

30 minutes. You leave with a clear answer: what we would build, how long it takes, what it costs, or a straight “this is not a fit”.

What we build

Working systems, not slide decks. Every project ends with a model running on your data, wired into your operation.

Forecasting & planning

Replace the spreadsheet and the gut feel with a model that forecasts from your history and updates itself. Better numbers for stock, staffing, pricing and cash.

Demand, sales and revenue forecasting
Stock and capacity planning
Pricing and promotion impact

Document & back-office automation

Invoices, contracts, claims, emails, tickets. AI reads them, extracts what matters, classifies, routes and answers, with a human in the loop where it counts.

Extraction and classification of documents
Assistants over your internal knowledge
Automated processing of repetitive workflows

Classification & risk scoring

Your team already knows what a good or bad case looks like. We train a model on their past decisions so every new case gets a score automatically. Your team reviews only the uncertain ones.

Credit, churn and default risk
Fraud in traffic, accounts and transactions
Anomalies in sensors, operations and billing

Need the data pipelines, dashboards or infrastructure that make this possible? We build those too, as part of the project.

How it works

One call, a written proposal the next day, a fixed-price build with a fixed end date. Maintenance if you want it. Pricing is shared after the call.

01
Call
30 minutes · free
You tell us the decision you want automated and what data you have. We tell you whether it is feasible, what we would build and roughly what it costs. No preparation needed.
02
Proposal
Within 24 hours
A written proposal: the data we need, the success criteria in plain numbers, the timeline and a fixed price. You sign off before anything is built.
03
Build
2–6 weeks · fixed price
We connect to your data, build the model and deliver the results where your team works: an API your systems call or a dashboard your people open. The build ends with a results presentation against the success criteria and a handover of everything we made.
04
Maintain
Monthly · optional · cancel anytime
We host, monitor and keep the system accurate: drift checks, retraining with new data, fixes and new features as your business changes. Or your team runs it; the choice is yours.

What you can count on

Fixed scope and fixed price, agreed in writing before we start.
Success criteria in plain numbers, agreed at kick-off, measured on delivery.
You own everything we deliver: code, models, tests and documentation.
Runs on your servers or on our cloud. Maintenance and hosting available if you want them.

Book a free 30-min call

Questions we get before the call

Is my problem a fit?: Good fit: a decision or a number your team produces repeatedly today (fraud / not fraud, how much to order, which category this document belongs to) and a few months of history where you can see what happened. If you have both, book the call. If you are not sure, book the call anyway: we will tell you in 30 minutes.
What data do you need?: Exports from the systems you already use: your database, your platform, spreadsheets, CSV files. Nothing has to be clean. Part of every project is a data audit where we tell you what is usable and what is missing. As a rule of thumb, a few thousand past cases with the outcome is enough to start.
What do I get at the end?: A results presentation with the numbers against the success criteria we agreed (for example: forecast error, share of cases classified correctly), and the working system: model, code, tests and documentation, delivered where your team works. Then you decide whether we maintain it or your team takes it over.
Who owns what you build?: You do. Code, models, tests and documentation are delivered to you. If you decide to run it yourself, or with another team, you can.
Where does it run, and who maintains it?: Either on your servers or on our cloud; we recommend once we know the data volume and where it lives. Maintenance, monitoring and retraining are available as a monthly service, cancel anytime. You can also take it in-house at any point.
How do you deliver this fast at a fixed price?: Our engineers spend their time on what matters: the spec, the success criteria and checking results on your data. AI agents under their supervision write and test the code. Less time typing, more time verifying, and a price that does not grow with the number of people on the project.

Which decision would you automate first?

Bring the problem. In 30 minutes you will know whether it is feasible, what we would build, how long it takes and roughly what it costs.