CALIBER™ — the process we use to move from uncertainty to an action plan signed off by leadership.
We don't start by implementing. We start by understanding. Each phase is a calibration point: we measure reality, contrast it with the vision, and only then decide. Judgement is built with data, not with trends or keynotes.
Every problem is quantified in euros. Every objective has a KPI, a baseline and a target. Every pilot has a continuation criterion agreed by leadership. If the number doesn't appear, we halt the phase until it does.
Not every process should be automated. Not every company is ready. If in any phase we find that investing in AI isn't a priority, we say so — and the consulting engagement stops. We charge for recommending what's right, not for executing what was sold.
Each phase closes with a concrete deliverable and a decision point. If the result doesn't hold up, we stop here — we don't move on to the next one. The consulting engagement is reversible by design.
Before measuring anything, we need to know where the company wants to go. We work directly with the CEO and the leadership team to map out the 3-year vision, the current revenue drivers, the competitive risks and the key decisions that need to be made. Without this map, any investment in AI is a blind bet.
Understand the company's direction and revenue model. If the result validates the hypothesis, we continue. If not, we stop here.
We visit the plant. We interview the heads of production, quality, maintenance, logistics and customer service. We time processes. We count reworks. We quantify downtime, errors and lost hours in euros. What is often perceived as a 'nuisance' becomes a number: this is the basis on which we decide what is worth automating.
Map the current reality (AS-IS) and quantify the problems. If the result validates the hypothesis, we continue. If not, we stop here.
Half of all AI projects fail because the data doesn't exist, isn't clean or can't be cross-referenced. In this phase, the technical team audits systems (ERP, MES, SCADA, spreadsheets), reviews data quality, identifies required integrations and assesses whether the internal team can maintain what gets built. If we find a serious blocker, we say so now — not after we've already signed with a vendor.
Validate infrastructure, data and technical capacity. If the result validates the hypothesis, we continue. If not, we stop here.
We turn the quantified problems into measurable objectives. Not 'improve maintenance' — but 'reduce unplanned downtime by 20% in 6 months with a monthly KPI'. We define who is responsible for each metric, how it is measured, what happens if it is not met, and what the minimum architecture is to get there. This is the phase where the CEO accepts or rejects each commitment.
Design the future (TO-BE): governance and architecture. If the result validates the hypothesis, we continue. If not, we stop here.
From the list of quantified problems, we select the 4–6 with the best impact/effort ratio and turn them into actionable use cases. Each case comes with a brief: functional description, required data, architecture, estimated ROI with a return horizon, risks and pilot plan. Leadership ranks them by priority — the roadmap is born here.
Prioritise use cases with estimated ROI and technical design. If the result validates the hypothesis, we continue. If not, we stop here.
We close the process with a document the CEO can sign and a team can execute. It includes the sequence of pilots for the next 12 months, a detailed budget (internal + external), continuation criteria for each pilot, a team training plan, and quarterly tracking metrics. The consulting engagement ends here — execution is done by the company, with or without our support.
Report, roadmap, budget and training plan. If the result validates the hypothesis, we continue. If not, we stop here.
Each phase includes a visit, interviews and real quantification of the problems in euros. Skipping measurement leads to generic recommendations. A phase closes when the number is there, not before.
The process is designed to end with a decision, not to drag on. Each phase has a deliverable that validates (or halts) the engagement. It never feels like a cost that never ends.
Each phase closes with a calibration point: if the result doesn't hold up, we stop. We don't drag the process along hoping things will improve. Judgement is built stage by stage.
An SME CEO can't dedicate day-to-day time to an endless internal process. Sessions are concentrated at the critical moments — the rest is resolved by email and with our team, without blocking operations.
The traditional model is designed to drag on. Caliber is designed to finish on time and with a decision made. This is the difference, row by row.
| CALIBER™ — FORECASTRAX | TRADITIONAL CONSULTING | |
|---|---|---|
| Starting point | Plant visit and quantification in €. | Generic workshop based on the client's own presentations. |
| Duration | 6 sequential phases with a fixed close date. | Open-ended process with billable extensions. |
| Final deliverable | Signed roadmap + fixed budget. | Strategic recommendations with no budget. |
| Continuation criterion | If a phase doesn't validate, we stop. | The process continues until the contracted hours run out. |
| Technical independence | No vendor agreements. We recommend what works. | Alliances with hyperscalers and software vendors. |
| Price | Fixed budget for the whole process. | Billed by the hour + retainers. |
| Team | Three senior partners on site. | One partner and a rotating team of juniors. |
| What happens next | The company has judgement. It can execute with whomever it wants. | Tied to the consultancy to implement what was recommended. |
Total: 6 meetings of 90' and 1 plant visit of 4h, spread across the six phases. The rest is our work. The leadership team gets involved at 3 critical moments — the rest is resolved by email.
Session with leadership. You share the vision, revenue drivers and pending decisions. No prior preparation needed.
We accompany the operations lead. You get on with business. At the end of the day, a brief joint read-out.
You see the AS-IS map with quantified problems. We validate priorities before continuing.
The IT team handles it. You receive a one-page summary of critical blockers, if any.
You decide KPIs and targets. This is where the leadership team signs off. It's the most important session in the process.
We present 6–8 cases. Leadership prioritises. You leave with the portfolio ranked by consensus.
Draft of the roadmap and budget. Review at your own pace, comment in writing.
Final validation of the budget and the schedule. Possible cuts or reallocations.
Handover of the executive report. Signing of the roadmap. Delivery of all materials.
If you wish, we start the first pilot. If not, the company has the judgement to decide it with any partner.
Metalworking sector · Baix Empordà · 2025
— CEO, industrial SME · 47 employees · €6.2M revenue
If you can't find yours, write to us directly at info@forecastrax.com. Reply within 48h, no obligation.
If the company is going through a liquidity crisis or a deep restructuring, AI consulting is not a priority. Also, if it does not yet have basic operational data (a working ERP, production records), you first need to put those pieces in place. We tell you this at the first meeting.
Fixed budget, no billable extra hours. The range depends on the size of the company and the number of plants to visit. For a typical industrial SME (40–120 employees, one plant), we're talking low five figures. We give you the exact budget after the first 30-minute call.
We say so clearly and the consulting engagement ends at phase 02. We only charge for the phases completed. This is the difference from a traditional consultancy: we have no incentive to recommend what won't work.
Yes. Phase 03 (Technical Validation) requires access to the main systems (ERP, MES, SCADA, data repositories). We work with your IT lead or with the external provider you have. We don't ask for sensitive credentials or copy data outside the company.
Yes. We have no partner agreements with hyperscalers, ERP, MES or AI vendors. Technical recommendations are made on technical merit, not on commission. If we recommend a tool, it's because it's the best fit for your case.
Yes, but it's not mandatory. The Implementation service is independent. Many clients execute the roadmap with their internal team, with another provider, or with us — always with a budget and KPIs agreed per pilot. The idea is that you reach the end with enough judgement to decide freely.
NDA signed before phase 01. All documents are shared in a private repository. The Forecastrax team that sees your data is the three founding partners — we don't subcontract the diagnosis. At the end of the process, you can request the destruction of working copies.