AI software pricing that makes sense

We quote fixed prices wherever possible. No billable-hour surprises, no scope creep without your sign-off. Below you'll find our three standard engagement types, what's included, and what to expect in terms of timeline and cost.

Choose a starting point

These packages cover the most common project shapes we see. If your needs sit between two tiers, or you want something entirely different, we'll put together a custom quote after a free 30-minute call.

Discovery and audit

$4,500 fixed

Best for companies that aren't sure where AI fits yet.

  • Half-day on-site or remote workshop with your team
  • Review of existing data sources and quality
  • Written report identifying two or three high-value opportunities
  • Rough cost and timeline estimates for each opportunity
  • One follow-up call to discuss next steps
Book a discovery

Build and deploy

$18,000–$45,000 fixed

A single AI feature or pipeline, from prototype to production.

  • Everything in Discovery, rolled into the project cost
  • Data preparation and feature engineering
  • Model training with documented accuracy benchmarks
  • Containerised deployment to your cloud (AWS, Azure, GCP)
  • API documentation and integration support
  • Three months of post-launch monitoring and bug fixes
  • Full code and model ownership transferred to you
Start a project

Ongoing partnership

$3,500 / month

Continuous improvement after your initial build.

  • Monthly model retraining on fresh data
  • Performance monitoring and drift detection
  • Up to 20 hours of development per month for new features
  • Priority response within 4 business hours
  • Quarterly review meeting with usage and ROI reporting
  • No lock-in: cancel with 30 days' notice
Discuss a retainer

Feature comparison

A side-by-side look at what's included in each tier. The Build and deploy package is where most clients start; the ongoing partnership is something you add once the first project is live and delivering value.

Feature Discovery Build and deploy Ongoing partnership
Workshop / needs analysis✓✓✓
Data quality audit✓✓✓
Opportunity report✓✓—
Model training and validation—✓✓
Production deployment—✓✓
API documentation—✓✓
Post-launch bug fixes—3 monthsOngoing
Monthly retraining——✓
Drift monitoring——✓
New feature development——Up to 20 hrs/mo
Quarterly ROI review——✓

What moves the price up or down

The range on the Build and deploy tier is wide because projects vary. Here are the factors that matter most.

Data preparation work on a laptop

Data readiness

If your data lives in a clean database with consistent formatting, we spend less time on preparation. If it's scattered across spreadsheets, PDFs and email threads, we need to build an ingestion pipeline first. That can add $3,000 to $8,000 depending on volume and variety.

Model complexity

A straightforward classifier or regression model is faster to train and validate than a multi-step pipeline involving computer vision, natural language processing, or reinforcement learning. Simple models sit at the lower end of the range. Multi-modal systems push toward the top.

Integration depth

Delivering a standalone API is simpler than embedding AI directly into your existing application, ERP or CRM. Deep integrations require more coordination with your dev team and more testing. We charge for this time, but we quote it upfront so there are no surprises.

Compliance requirements

Some industries need audit trails, explainability reports, or on-premises deployment to meet regulatory obligations. We build these in when required, and the extra engineering adds to the total. We'll flag these costs in the discovery report so you can decide what's worth doing.

Common questions about pricing

No. The first call is free and usually lasts about 30 minutes. We use it to understand your situation and decide whether a Discovery engagement or a direct Build makes more sense. If we don't think AI is the right tool for your problem, we'll tell you that too.
Yes. If you proceed to a Build and deploy engagement within 60 days of the Discovery report, we subtract the $4,500 from the project total. The Discovery work feeds directly into the build, so nothing is wasted.
Our fixed-price quotes include a contingency buffer. If something genuinely outside the agreed scope comes up, we'll present a change request with a clear cost before doing any extra work. You approve or decline. We've never billed a client for work they didn't sign off on.
Yes, completely. At the end of a Build project, you receive the source code, the trained model weights, all data pipelines, and deployment scripts. We don't retain any proprietary claim. If you want to hire an internal team to maintain it later, everything transfers cleanly.
No. It runs month to month after an initial three-month commitment. You can cancel with 30 days' written notice. We find most clients stay because the monthly retraining and monitoring keeps their models accurate as data patterns shift, but we don't force anyone into a long contract.
Absolutely. About half our current clients are in other states. Discovery workshops can be done remotely via video call, and all deployment work happens through your cloud provider. We travel for on-site workshops when needed, with travel costs quoted separately.