Nestlé
Nestlé — Everyday AI
Half-day Copilot workshops delivered across teams, with a personalized AI Academy now being scoped as the next step.
See full storyB2B Offering · Ironhack for BusinessCompanies already use AI.
Companies already use AI.
What they don't have is order.
An AI adoption model with three levels of depth: audit, task matrix, standards and governance, built from real delivery experience.
Short coursesAI ConsultingAI School
PRODUCT GUIDE · B2B WEBSITE · JULY 2026
IRONHACK COMMUNITY EVENT · TECH TALKS FORMATThe real market
The question is no longer "are we using AI?", it's "who's using it, how, and with what control?"
79%
of Spanish companies with more than 50 employees already have at least one AI tool deployed.
IDC, AI Adoption Gap Spain 202623%
report a measurable impact. The rest: scattered use, no standards, and no measured return.
IDC, AI Adoption Gap Spain 20269/10
were already using AI on their own (Claude, ChatGPT, Lovable), each in their own way, with no control or shared criteria.
Real discovery, 5-company holding group, Jun 2026The problem isn't a lack of AI, it's a lack of control: scattered use, no data governance, and no measured impact. The value proposition is turning that use into a governed, measurable model, with shared criteria across the whole workforce, data under control, and return quantified by task.
And there's a regulatory clock ticking: art. 4 of the AI Act requires AI literacy training for the entire workforce since Feb 2025, with penalties in force from Aug 2026. Governance stops being optional and becomes a requirement with a deadline.
The thesis
It's not a course catalog: it's an adoption model with three levels of depth
01 · AUDIT
Who uses what
In which tasks, with what risk.
02 · MAP
The 2x2 matrix
The whole company, task by task.
03 · STANDARDIZE
Role-based training
Sets the standard for working with AI.
04 · GOVERN
Usage policy
Approved tools, AI Act.
05 · MEASURE
On-the-job KPIs
Real application, not attendance.
Level 1 · Activation
Short sessions
The model in a half-day session on their own stack. The team leaves with improvements they can apply right away. Scalable to the whole workforce.
€4–6k / sessionLevel 2 · AI Consulting
Assessment
The full model: task-by-task map, roadmap, governance, and rollout of the first improvements.
Lite €8–12k · group €30–45kLevel 3 · AI School
Implementation
The model plus the training that installs the role-based standard and leaves internal capacity, with in-house trainers.
From €25k, assessment includedThe common base is the task-by-task assessment. These aren't three products, it's the same model at different depths, so the client can start with the basics and expand without redoing anything. And it's measured by outcome: processes in production, hours freed up, and AI Act compliance.
Level 1 · Activationfrom €4–6k / session
Courses on Claude, Copilot and ChatGPT, with results you can apply right away
Short sessions on the client's own stack. The first level of the offering, scalable to the whole workforce.
Three formats
- Team productivity, for managers: email, reports, analysis, day-to-day workflows. Can run by department.
- Technical / code, for development teams: generating, refactoring, testing, pair-programming.
- Executive session, for the leadership committee, the one that connects to the assessment.
- Flexible format, half-day sessions. Claude as the flagship tool, Copilot or ChatGPT depending on the stack.
How it works
- 10-minute pre-work: a survey that personalizes the session and yields the first audit data.
- Proven structure: framework, hands-on, prompting framework, building on their own use case.
- Closes with commitments to apply what they learned, signed in the room, and measurable afterward.
What the client walks away with
- Editable slides, prompt templates, worked exercises.
- A replicable demo agent, built on their own use case.
- Add-on: recorded micro-lessons by profile, to scale to the rest of the workforce.
Format proven via IronhackNestlé — Everyday AI: half-day workshops on Copilot.AstraZeneca — steering committee + 9 sessions + 8 recorded micro-lessons, for 163 people across 5 profiles.
Level 2 · AI ConsultingLite €8–12k · group €30–45k
The assessment that turns scattered AI use into a prioritized plan
Phase by phase, with estimated savings per task and a go / no-go decision at the end of each one.
What we do
- We audit current use: which AI tools are already in circulation, including unauthorized ones, with what data and what risk.
- We map the whole organization onto the matrix, task by task, with a real census and interviews.
- We roll out the prioritized solution in phases, with our own trainers and technical staff, and measure real on-the-job adoption.
Formats and scope
- Lite Assessment, 3 weeks: survey, matrix, and first improvements.
- Full group assessment, 6 to 8 weeks: census plus 10 to 12 interviews.
- Scope and price fixed per phase. Benefit is measured in hours freed per process and its equivalent in euros, never a generic global figure.
What you walk away with
- A complete automation map, task by task.
- 3 to 5 prioritized quick wins, with estimated hours and euros saved.
- A business case with their own numbers, and data and AI governance: usage policy, approved tools, and a register, exactly what the AI Act already requires.
Real-world case, 2026A €25,000 program signed with a food-industry holding group, with its €8,000 Lite Assessment included. Honest expectations: a managed census with 55 to 70% response, not a promise of 100%.
A CENSUS AND MATRIX WORKING SESSION, ON SITE WITH THE CLIENTAI Consulting · the 2x2 matrix
Where you stand today and how far you can go, task by task, on a single page
The assessment maps each task's real starting point and target state, using the client's own data. Illustrative example.
Operational · day to dayStrategic · planning
Order entryCustomer supportSales follow-upProposal draftingKey negotiation
Customer segmentationSales forecastingCompetitive analysisPricing policyProduct definition
Billing and collectionsAccount reconciliationDocument managementPayroll and on/offboardingProcurement
Reporting and BIBudget controlTalent selectionGDPR / AI Act complianceInvestment decisions
Core (top row) and support (bottom row), read from operational to strategic.
The traffic light
AI executes, human reviews
AI accelerates, human validates
Human leads, AI assists
- It also maps shadow AI: what each team uses without oversight.
- Built from real census data, not a generic questionnaire.
- The green and amber tasks become the first improvements, and feed AI School's learning paths.
The website's edge
A calculator that puts savings in euros, and makes the course nearly free
Enter your numbers and discover, in real time, how much time your team frees up with AI done right, and how much FUNDAE gives back.
Realization factor, conservative
Hours freed · year 14,992 hemployees × hours/week × 52 × realization factor
Estimated savings · year 1€139,776payback in 1 month (Lite Assessment, €10,000)
Estimated FUNDAE credit€3,750on €5,000 of training budget, indicative bracket 75%
A directional estimate for sales conversations, using a conservative realization factor. It doesn't replace the assessment or current FUNDAE regulations, which depend on headcount and the credits the company has already used.
Level 3 · AI Schoolfrom €25k
The in-house academy that sets the standard for working with AI
The model plus role-based training, with internal capacity to sustain it, assessment included.
Architecture
- Group level, power users: cohorts by role, more than 50% hands-on, real automations live in production before finishing.
- Leadership level, 1:1: a personal agent per executive, connected to their data and email. They decide, the agent executes.
- Train-the-Trainer: an internal network of trainers to sustain the program autonomously.
The differentiator
- Learning paths come from the client's own discovery: real use cases with an owner, a validated prompt, and a KPI. Never a canned catalog.
- Governance as a module: usage policy, approved tools, evidence of AI Act art. 4 compliance.
- Sign-off KPIs at 3 layers: individual commitments, per-case KPIs, business results. Never attendance or hours delivered.
Why now
- FUNDAE-subsidized: AI is a priority sector in 2026.
- AI Act art. 4: mandatory AI literacy since Feb 2025, penalties in force from Aug 2026. A budget with a deadline.
- If it follows Consulting, it starts with the map already done: zero weeks of duplicated assessment.
Base realInstructor for Nestlé's Everyday AI program (on Copilot). A 13-week AI Academy currently running at a food-industry holding group (Claude + Lovable): 25 power users plus 1:1 leadership coaching, a catalog of 30 to 40 use cases sourced from discovery.
AI School · learning paths
From how each role works today to how it will work with AI done right
One path per maturity level, from basic use to production. Each cell becomes a use-case card: task, tool, savings, KPI, and owner.
| Role | Foundationaluse AI well | Intermediatebuild with AI | Advancedoperate in production |
|---|---|---|---|
| Sales | Prompting for proposals, email, and visit prep. | Proposal assistant with CRM and catalog context. | Semi-automated proposal pipeline, with human review. |
| Finance / Admin | Summaries, reconciliations, and assisted email. | Semi-automated monthly reporting straight from source data. | Closings with agents connected to the ERP, with validation. |
| Operations | Assisted documentation, reports, and incident management. | Automated workflows (Make / n8n) on their own processes. | Automations in production, with QA and an assigned owner. |
| Leadership | Daily briefing and assisted document analysis. | Conversational dashboard over their own data. | Personal agent connected to data and email; they decide, the agent executes. |
Why Ironhack
A proven, repeatable model, ready to scale across accounts
01
Proven, standardized method
Census, matrix, and use-case catalog already built. Same quality on every account, nothing reinvented.
02
Network of trainers and technical staff
Scales without depending on one person. Availability and delivery guaranteed.
03
Quality QA on every deliverable
Every report and proposal is checked before it reaches the client. Zero surprises in front of the account.
04
Multi-stack
Claude, Copilot, ChatGPT. Fits any account, whatever their ecosystem.
Reference investment
Activation · Claude courses, short session€4–6k
Lite Assessment · 3 weeks€8–12k
AI School · 12–13 weeks, ~30 people€25k+
Full group assessment€30–45k
Internal reference based on real 2026 projects. Final pricing is set by Ironhack based on account and scope.
Client proof
Real stories with Ironhack
The same activation, consulting and school model, already delivered.
AstraZeneca
AstraZeneca — Data-Driven
A global sensitization session plus 30 internal ambassadors trained in a 40-hour program, with role-based rollout next.
See full storySantander
Santander — talent pipeline
Java bootcamps since 2021 across 5 countries, plus OpenBank's #WomenInTech cohort: 37 and 16 hires respectively.
See full storyCaixaBank
CaixaBank — Data & AI School
A 4-year, single-partner tech school for 6,972 employees across AI, Business Analytics and Digital Transformation.
See full storyLet's start with the assessment
The same model, at whatever depth your company needs today
Nothing to redo if you want to go further later. Let's talk about your case.