We write your HR-AI strategy in a month. Then we build only what it calls for.
Strategy first, on one fixed fee. Then a working proof on test data. Then production, inside your environment. Each step is scoped on its own, on a fixed fee, and each ends with a written go / no-go.
Thirty minutes. We start with your numbers.
- 01
Strategy
A sequenced roadmap and business case for your function
- 02
Proof
A working system on test data your team can use
- 03
Build
A production system in your environment, yours to run
Ongoing support retainer · optional · after the strategy
What the strategy month answers.
Five questions, worked through for your function.
These are the same five questions on our homepage. The month is the work of answering them for you.
- 01 Which number you already report that you'd most like to move
- 02 Which of those AI can actually help with
- 03 In what order
- 04 What stays human, no matter what
- 05 Whose name goes on each number
The three steps: what each one is, what it produces, and what it asks of you.
-
01 Strategy
- Duration
- About a month
- Commercials
- One fixed fee for the month, quoted after the first call and scaled to the size and complexity of your function
- From your side
- You, your HRIS or people-analytics owner, a handful of working sessions of about two hours, plus one data pull between them
- Ends with
- A written go / no-go on whether anything is worth building
We start from the numbers you already report on and which of them you'd most like to move: attrition in the roles you're most worried about, time-to-fill, the hours your team spends on repeat questions from employees. We also take an inventory of any AI tools already in use in HR.
Every idea, ours and yours, goes through the same four checks: what it's worth if it works, what it costs if it's wrong, what data it needs, and whether people will use it. The ones that survive get sequenced, safest and most provable first.
Before anything is built, we write down what AI will not decide in your function, and you name an owner on your side for every number.
The roadmap and business case are yours whether or not you go further.
- A prioritized, sequenced roadmap for your function
- A business case for each use case, in the numbers your CFO tracks
- The guardrails, written down, for your security team and works council
- An inventory of AI already in use in HR
- A one-page version for the exec meeting
What you leave the step with
-
02 Proof
- Duration
- A few weeks per use case
- Commercials
- Fixed fee per use case, quoted from the roadmap
- Where it runs
- Your choice: our sandbox on synthetic data, or a sandbox in your own cloud tenant on synthetic or de-identified data your security team approves
- Ends with
- A written go / no-go on production
For each use case you take forward, we build a working version in one of two places, your choice. The fastest is our sandbox on synthetic data: no IT dependency, no procurement, no live records. Or, for teams who want the proof inside their walls from day one, a sandbox in your own cloud tenant, set up with your security team, on synthetic data or a de-identified extract they approve. Your team uses it for a couple of weeks and we measure how it behaves.
The go / no-go is written against what your team saw in use: accuracy, the failure cases, cost per query, and whether people reached for it.
- A working system your team has used
- Accuracy, failure cases, and cost per query on data shaped like yours
- A written go / no-go recommendation
- The sandbox torn down, or handed to your team, at the end
What you leave the step with
-
03 Build
- Duration
- Typically a few weeks to a couple of months per use case, scoped after the proof
- Commercials
- Fixed fee for the agreed scope, quoted after the proof
- Where it runs
- Inside your environment, under your security team's controls
- After handover
- A support period after handover is included, scoped with the build. Hosting and model usage are billed to you by your own providers.
We build to your security team's guidelines and bring the architecture document and data-flow diagram to the first review meeting.
How people will use it, and which number tells you it's working, is agreed before anything ships. Code, documentation, and runbooks transfer to you at handover.
If your own team is going to build it, the roadmap and proof are written so they can, and we'll hand over cleanly.
- A production system in your environment, under your controls
- A usage and outcome report your team can pull without us
- Onboarding for the people who'll use it
- Code, documentation, and runbooks, transferred at handover
What you leave the step with
Ongoing support
For teams who want us to stay: owning the roadmap, sequencing the next use cases, and sitting in the security reviews as the program grows. We call it the Program Alliance. It's a retainer, offered only once we've done the strategy with you, and it isn't needed to keep a delivered system running.
- Format
- Retainer
- Eligibility
- After the strategy step
- Term
- Month-to-month
The facts a CIO usually asks for.
- Sandbox
- Ours on synthetic data, or one in your cloud tenant set up with your security team on synthetic or de-identified data. Your choice.
- Model
- The model runs where you choose, on the provider your security team approves, with the provider's no-training and retention terms in writing.
- Access
- Through your identity and your devices. We hold no production credentials after handover.
- Documents
- Architecture document and data-flow diagram before the first review. A written list of what AI will not decide.
- Ownership
- Code, prompts, evaluation sets, documentation, and runbooks are yours. No Culturro platform dependency.
- Certifications
- We hold none and claim none. The system runs under your controls and inherits your posture; what you assess is our process, not our data center.
What we won't build.
Some things stay human.
We don't build systems that make autonomous decisions about people: hiring, pay, promotion, discipline, termination. A human decides, always, and we design so that's structurally true, not just policy. AI may inform, surface, rank, summarize, or draft; it never decides. Anything a works council or regulator would treat as high-risk gets that human-decides design in writing in the strategy step.
Start with the conversation.
Thirty minutes on your function and your numbers. We'll go through the five questions above with you and tell you what we'd do first.
Bring the numbers you already report on.