Nucleoo.ai Book a call

Scattered systems. One clear next move.

Nucleoo connects the systems your teams already use, shows what is happening across the whole business and turns it into decisions and actions your people approve.

Book a discovery call See how it works

Marketing, sales, operations, finance. Each sees part of the picture.

Marketing creates demand, sales turns it into revenue, operations delivers and finance keeps score. Each works from its own systems, data and priorities, so decisions that affect the whole business get made without the whole picture.

The more systems you run, the more happens between them that no one sees.

The answers sit between your systems.

A monthly report follows a handful of measures, once a month. Nucleoo reads every variable your systems hold, every day, including data that conventional business intelligence was never built to join.

Nucleoo's read: online returns wait in the warehouse while the same sizes run out in stores, so store sales fall and margin follows.

Proposed: route online returns straight to the nearest store that is short of that size, instead of back to the warehouse. Waiting for your approval.

Illustrative: a fashion retailer with stores and an online shop. Seven measures from tools it already runs, and how Nucleoo connects them. Point at a measure, or tap it, to see what it raises, lowers or follows.

Every variable, not a handful

Sales, margin and costs, and everything your tools already record: the CRM, the ERP, invoicing and payments, ad platforms, web analytics, the support desk, the online shop and the spreadsheets in between. Nucleoo connects systems that were never joined before and reads them together, every day.

What no single report shows

Nucleoo tests how variables move together, which lead and which follow, and infers the drivers behind a result, including second-order effects that stay invisible inside any one system.

A strategy built on evidence

From those facts, Nucleoo proposes the strategy: what to change, where and in what order, with the evidence behind every recommendation. It does the reasoning; your people make the call.

A good fit when

  • Data, teams or locations work in silos.
  • Reporting is manual, or depends on one person to make sense of it.
  • Decisions need to come sooner, and with better context.

One cycle, run continuously.

Nucleoo works as a layer of intelligence on top of the systems you already run: connect, understand, decide, act, then measure and learn, and round again. Every action creates feedback. Every result informs what comes next.

Why are sales growing while profit is falling?

Illustrative example. No real client data.

  1. The starting point

    Your answer is already in your systems, scattered across them.

    Advertising spend in the marketing tools, opportunities in the CRM, orders and capacity in the ERP, costs and payments in finance, complaints in customer service.

  2. 1Connect

    Bring together what your systems already know.

    Nucleoo reads the tools already in place and lines up their data on one timeline, so you can finally read each department's figures against the others.

  3. 2Understand

    See what is happening, and why.

    Sales are rising, but demand is shifting towards lower-margin services and putting pressure on two teams. That is why profit falls.

  4. 3Decide

    A clear recommendation, with its priority.

    Shift part of the marketing spend towards higher-margin services that have spare capacity, and prioritise the order backlog by profitability.

  5. 4Act

    Nucleoo prepares the work; a person approves it.

    A proposed campaign adjustment for marketing, an order-priority list for operations and a cash-flow forecast for finance, each ready for its owner to review. Nucleoo recommends by default; automation runs only within permissions you approve.

  6. 5Measure

    Track the outcome against the starting point.

    Nucleoo follows margin, team workload and cash flow against the baseline for the weeks after the change. Only approved actions run, so Nucleoo measures only those.

  7. 6Learn

    The result shapes the next round.

    What worked, and what did not, refines the next analyses, decision criteria and priorities. Learning here means your business gets better informed; it does not mean retraining AI models.

The same cycle, different questions.

Success story: IAhoy runs its business on a weekly evidence loop.

IAhoy is Nucleoo's own AI assistant for the self-employed in Spain (autónomos), working inside WhatsApp. IAhoy connected its real data to Nucleoo and turned what it found into a change every week, with the team it already had: in marketing, product, pricing and sales.

€90€4

Cost per registration in marketing, about 20 times cheaper. Verified in the Meta Ads API.

Which message brings the right people in, and at what cost?

Understand
Reads ad results with usage statistics, after taking out test traffic and an ad pixel that counted sign-ups twice. Invoicing turned out to be only part of the job, and much of it happened in the evening.
Decide
Sell the whole admin in the WhatsApp people already use, and give them their evening back.
Act
New ad scripts every week, one flight at a time, with the winning scripts reused on new faces.
Measure and learn
Cost per registration every week, on the same definitions. A filmed presenter ad and a retargeting test showed no gain; Nucleoo reported both and the team dropped them.

What is breaking, and what should we build next?

Understand
Reads how people use IAhoy as it happens: bugs as they appear, user flows nobody had designed for, and the features customers keep asking for.
Decide
Fix the blockers first and build the most requested features, matched against the open tickets.
Act
Reproducible bug reports for the developers, and the most requested features shipped to production first. Marketing then talks about what customers were asking for.
Measure and learn
Activation and finished documents every week, on the same definitions. Within a week, Nucleoo caught a release that hurt activation, even though its error rate looked better.

Where should the free plan end?

Understand
Reads real usage per user against the plans, and who pays in Stripe, without counting fully discounted coupons as paying customers.
Decide
Set the limits where real use meets them, so upgrading makes sense and the paid plans stay fair.
Act
Simulates how many users each proposed limit would affect before anything changes, then sets the limits.
Measure and learn
Follows upgrades and usage after the change, on the same definitions, and moves a limit that bites in the wrong place.

What does an accounting firm need to hear?

Understand
Reads recorded sales calls with gestorías (accounting firms) for intent, objections and where a call gains or loses time.
Decide
Which leads to follow up first, by the intent they showed, and what to change in the script.
Act
An updated call script and playbook, and each lead's own words captured in the CRM as qualitative data.
Measure and learn
How calls run and how leads move through the pipeline after each script change, and the next questions to test.

The evidence was always there. Now someone reads it every week.

Four more questions, by kind of business

Illustrative examples. No real client data.

Why are margins falling at some of our stores?

Understand
Compares store-level sales, discounts, costs and stock, and finds a category where discounts outside the planned promotion coincide with lower margins.
Decide
Review those discounts first at the two stores with the greatest potential impact.
Act
Drafts a task for each store manager with the proposed change, for approval.
Measure and learn
Compares margins and stock turnover with baseline and targets. If the results do not support the hypothesis, that finding shapes the next analysis.

Why is one location less profitable than the others?

Understand
Combines recorded hours, payroll costs, invoiced projects and operational issues by location. Non-billable time has been rising at one site for two months.
Decide
Review team allocation and set a threshold for non-billable hours at each site.
Act
A weekly alert for the operations manager and a task to review resource planning.
Measure and learn
Monitors billable hours after the change and refines the threshold over time.

Why is cash tight at month end when we're invoicing more?

Understand
Connects invoicing, outstanding receivables, customer and supplier payment terms and fixed costs. The gap between invoicing and collection is widening, several large customers pay late and supplier payments cluster on the same days.
Decide
Follow up the most overdue, highest-value receivables first, stagger supplier payments and review payment terms in new quotes.
Act
Payment reminders for approval, a proposed payment schedule for finance and a note to sales on payment terms, pending review.
Measure and learn
Tracks days to collect, overdue balance and month-end cash against forecast, and adjusts the next priorities.

Which customers are at risk of not buying again?

Understand
Analyses orders, purchase frequency, delivery issues and reviews, and finds customer groups whose purchasing drops after a delivery problem.
Decide
Prioritise higher-value customers, with tailored messaging for each group.
Act
Prepares audience segments and email drafts in the marketing platform, for review before sending.
Measure and learn
Compares repeat purchases among contacted customers with a similar group that was not contacted, and uses the outcome in future campaigns.

Retail and multi-store, services with several teams or locations, e-commerce and distribution: every industry asks different questions and measures success differently. If yours isn't listed, the questions will differ but our approach won't.

Start with one question. See first results in days.

We begin with one focused, measurable use case and expand from there. We can build and test a first working prototype in days, then add connections and capabilities once it proves its value.

  1. Discovery and assessment

    A focused starting point

    We review your systems and workflows, choose a high-value first use case, set baseline measures and agree goals and expected return.

  2. Implementation

    Built for your business

    We connect the relevant systems, configure the solution for the agreed use case, check data quality, permissions and workflows, and launch something you can use and measure.

  3. Continuous optimisation

    A monthly service

    We monitor performance, measure outcomes against the agreed goals and refine analyses, criteria and workflows. An evolving capability, not a maintenance contract.

One embedded team, two rhythms, and your people from day one.

Rapid discovery
Cycles in days
Explores what is possible, builds prototypes, tests and iterates.
Can we make this work?
Production engineering
Longer cycles
Integrates and secures data and access, then deploys and runs it reliably at scale.
Can we trust this in production?
Your team
From day one
Observes, builds alongside us, operates and owns the result.
The knowledge stays in your business.

Every system connected. Every permission agreed.

Together with you, we define who can see what, which actions need approval and how data is handled in your deployment. Five questions, answered before anything is connected.

Continuous improvement never means unrestricted autonomy: outcomes are evaluated within the rules and permissions you set.

  1. What can Nucleoo do on its own, and what needs approval?

    Nucleoo recommends. Automation runs only within permissions you approve, a person signs off sensitive actions before they run, and Nucleoo logs activity as agreed in the deployment scope.

  2. Who can access which information?

    We agree with you which data sources and tools Nucleoo can access, and how the architecture keeps your data separate.

  3. Which certifications apply, and what do they cover?

    Nucleoo is certified to ISO 9001 and ISO/IEC 27001. ISO/IEC 42001 certification for AI management is in progress. We share each certificate's issuer and scope, and keep Nucleoo's own certifications separate from those of third-party providers.

  4. Where is data stored, processed and sent?

    For each deployment, we document where data is processed and stored, what is copied and which providers are involved.

  5. What happens to the data if you leave?

    We document export and deletion, including any retained backups, before you start.

Works with what you have

Nucleoo integrates with your existing tools wherever their data is accessible: ERP, CRM, point of sale, invoicing, inventory, online stores, marketing and support platforms, and the spreadsheets that hold the rest. The assessment confirms which connections are possible and what each involves. There is no new platform to adopt.

Certified and compliant

Nucleoo is certified to ISO 9001 for quality management and ISO/IEC 27001 for information security, and complies with the General Data Protection Regulation (GDPR) and the EU AI Act. ISO/IEC 42001 certification for AI management is in progress.

  • ISO 9001Quality management

  • ISO/IEC 27001Information security

  • ISO/IEC 42001AI managementComing soon

  • GDPRData protection

  • EU AI ActAI regulation

15 years building technology for businesses.

Engineering, integration, data and automation. Nucleoo.ai brings that experience, our reusable components for integration, automation and control, and leading AI models to your business.

Earlier work by the Nucleoo team, published on nucleoo.com.

All case studies on nucleoo.com

Questions we're asked first.

Will we need to replace our existing tools?

Not necessarily. Nucleoo integrates with your existing tools where their data is accessible. The assessment confirms which integrations are possible and what is involved.

Where do we begin?

With one focused use case and clear success measures, chosen together during the assessment, with a baseline and agreed target outcomes.

How long does implementation take?

It depends on the number and complexity of your data sources. A first working prototype takes days; the assessment gives an estimate for the full implementation.

Do we need to clean up our data first?

Not necessarily. Nucleoo helps find inconsistencies, flag quality issues and apply agreed data-quality rules. It does not invent data to fill gaps in your sources.

How does Nucleoo use our data?

To support your business's own analyses and recommendations, within agreed access controls. The proposal sets out where data is processed, what is retained and which providers are involved.

Will our data be used to train AI models?

The providers and configurations used in each deployment are documented, including what data the models receive and the training and retention terms that apply.

Can Nucleoo take action without approval?

By default it recommends rather than acts. Automations run only once you authorise them, and sensitive actions need a person's approval.

What if Nucleoo gets something wrong?

You can check every recommendation before anything happens, a person reviews sensitive operations and Nucleoo measures the results. When an assumption proves wrong, Nucleoo uses that finding in later analysis.

How does it improve over time?

By measuring outcomes, comparing them with targets and revisiting the assumptions behind each recommendation. Results inform the context, priorities and decision criteria of future analyses, within agreed controls.

How is Nucleoo priced?

It depends on the use case and scope. After a first conversation we define the assessment, and a tailored proposal covers implementation and ongoing optimisation.

Tell us how your business works. We'll find the first move.

In a discovery call we look at three things with you. From there, we identify a practical first opportunity to apply AI, with a baseline to measure it against.

  • The systems you run, and where the data sits
  • The decision you most want to get right
  • A first use case, and how it will be measured

We use these details only to reply to you and arrange the call.