How we work

Sprint, unlock, then capability waves

An AI-Native Sprint sets the baseline and the plan. Unlock changes only what the prioritized capabilities require. Capability waves then ship production AI, each wave reusing the foundations of the last.

The method

  1. 01AI-Native Sprint
  2. 02Unlock
  3. 03AI Capability Waves

Fixed scope, fixed duration, a written definition of done.

  1. 01
  2. 02
  3. 03
01

AI-Native Sprint

Establish the product, workflow and architecture baseline, prioritize the AI capability portfolio and define the first wave.

  • Product, workflow and architecture baseline with the people who built it
  • AI capability portfolio prioritized on value, feasibility and payback
  • Security, tenancy and permission model established up front
  • The first capability wave defined, scoped and sequenced

How we sell it

One buying journey: Review, Sprint, Waves.

Datics has run as a software factory for years: repeatable delivery pods, fixed scopes and codebases we did not write. Every AI engagement here has a name, a duration and a defined end state.

  • Fixed scope and fixed duration on everything before ongoing waves.
  • The same delivery pod shape on every engagement: BA, PM, engineers, QA.
  • Written definition of done, agreed before we start.

Indicative ranges. Final price is fixed in writing before any work starts.

01Free

AI Product Review

30 minutes · No prep required · No pitch deck

We walk your application with you, name the highest-value AI capability and tell you what has to change in the product before it can ship.

What you receive
  • Priority workflow opportunities in your product
  • Architecture observations from walking the application
  • Technical constraints that shape what is possible
  • An honest read on AI readiness
  • Major risks, including security, tenancy and governance
  • Recommended next steps and what to build first

Done means

A written recommendation on what to build first and why.

Product and engineering leaders deciding where AI belongs first.

Book a 30-minute product review
02$12k - $18k

AI-Native Sprint

Indicative · 3 to 4 weeks

The product, workflow and architecture baseline, a prioritized AI capability portfolio, the security and tenancy model, the unlock work each capability requires and a defined first wave.

What you receive
  • Product, workflow and architecture baseline
  • Prioritized AI capability portfolio
  • Unlock work each capability requires
  • Integration map
  • Security and tenancy model
  • Product flows and interaction design
  • Implementation sequence
  • First capability wave, defined and sequenced

Done means

A sequenced plan you can fund, staff and defend in a security review.

Teams with an AI mandate who need a plan their board and their architects both accept.

Scope a sprint
Most common start
03$35k - $50k

AI Capability Wave

Indicative · 8 weeks · In production

Unlock plus one capability shipped inside the product you already run: context, tools, permissions, tenant boundaries, approval gates, audit trail, evaluation and telemetry.

How it works
  • We work inside your existing product, architecture and delivery environment
  • Prioritized capabilities taken into production for real users
  • Security, tenancy and audit designed in, not retrofitted
  • Evaluation and observability left running behind the capability
  • A reusable foundation for the next capability

Done means

One AI capability your customers use inside the product, plus reusable foundations.

Product teams who have decided what to build and need it live with real users.

Scope a capability wave
04From $18k / month

Ongoing Waves

Retainer · Rolling · Minimum 3 months

A standing delivery pod running wave after wave against your roadmap: business analyst, product manager, engineers and QA, working inside your repos, rituals and release process.

What the pod includes
  • 1 business analyst and 1 product manager
  • 2 to 4 AI product engineers
  • 1 QA engineer with evaluation ownership
  • Fortnightly shipped increments into your environment
  • Shared harness, tooling and observability across every capability

Done means

Predictable AI throughput without hiring a new team.

Companies treating AI as a permanent product line rather than a project.

Scope ongoing waves

Prices are indicative ranges for scoping conversations. Final scope, duration and fee are agreed per engagement.

Solutions

Named solutions live inside a wave

These are things we build during a capability wave, not separate processes to buy.

  • Copilot Launch

    An embedded copilot inside an existing workflow, on your APIs and permission model.

  • Agent Harness Foundation

    The reusable layer under every later capability: tools, policy, memory, evaluation, observability.

  • Feature discovery

    A natural-language layer that takes users to functionality you already shipped.

  • Supervised agent UX

    Visible steps, interrupt and an approval gate before the first irreversible action.

  • Cloud and AI workload modernization

    The infrastructure work a production AI capability depends on, across AWS, Azure and Google Cloud.

What you can expect

Working inside software you did not write is a discipline

We take responsibility for the product as it is: the integrations, the permission model, the accumulated business logic and the customers who depend on all of it.

We change what needs changing

No speculative rewrites. Architecture work is scoped to the AI capability it enables.

We ship into production

The measure of success is a capability your customers use, not a demo environment.

We leave a foundation

Tools, context, evaluation and governance your own engineers can build on.

Start with one capability that matters.

Bring us the outcome you want and the product you already have. We will tell you what step one looks like.

Book a 30-minute product review