01 — Ownership
The software is yours
Code, data and AI models are your property. Forever. No subscription fee holding your processes hostage.
Front page · Business Case · 2026 Edition
A business that costs €450,000 a year today can run on €48,000.
We install an AI engine inside your company. In twelve months, from zero automation to a semi-autonomous business — the software, the data and the models remain your property.
Most companies treat artificial intelligence as an experiment: a chatbot here, an automation there. This prospectus makes a different case — AI as the company's operating system, connected to every process, every piece of data, every decision. Not in five years. In twelve months.
Editorial · The thesis in brief
AI is not a project. It's an operating system.
Most companies treat AI as an isolated experiment. We think of it as the company's nervous system — connected to every process, every piece of data, every decision. Not in five years: in twelve months. Every step generates measurable value before the next one begins.
02 — Infrastructure
The hardware is ready
On-premise or dedicated cloud. Your data runs wherever you decide, behind your own security perimeter.
03 — Method
The journey is gradual
Three Eras, twelve months. Each phase can be purchased on its own and delivers a return before funding the next.
The journey · Timeline
Three Eras. Twelve months. An engine that grows with you.
A three-phase roadmap. Each Era switches on new capabilities and a measurable return, before opening the next.
-
01
Ignition
Months 1–2 · live in 48 hours
Switch on the first 12 core automations and prove immediate value.
- Product images and videos, social content
- Orders from voice notes, customer-specific price lists
- Pre-visit client briefs, site reports from audio
- Expenses from receipt photos, invoices, documents
- Recruiting, contracts, 24/7 customer support
Small business · 5–15 employees
Estimated ROI≈ €65,000/yr1–2 FTEs
-
02
Orbit
Months 3–6 · answers in 8 seconds
Collaborative intelligence and full visibility over the company. Nine modules.
- Company Knowledge Base, Email Intelligence
- Swarm Intelligence — multiple AI Workers collaborating
- Meeting Intelligence (Meet / Zoom / Teams)
- Computer Use (Playwright), Dev Pipeline
- Metric Intelligence Engine — 60+ auto-discovered KPIs
Mid-sized company · 15–50 employees
Estimated ROI≈ €180,000/yr3–5 FTEs
-
03
ImpresaPilot
Months 6–12 · operational autonomy
The company runs even when the CEO is away. Over 30 modules.
- Autonomous Decision Engine, 24/7
- Cash Flow Forecasting, Customer Health Score
- Fine-tuned AI models (LLaMA / Qwen) on your documents
- 30+ coordinated AI Workers, autonomous pricing
- Dedicated on-premise GPU · “My Day” brief at 7:00 am
Structured company · 50–200+ employees
Estimated ROI≈ €400,000/yr8–12 FTEs
Dispatches from the field · Seven reports
Seven real cases. One engine.
How ImpresaPilot works across departments, told from the point of view of the people who use it every day.
Sales
The voice note that becomes an order
A sales rep sends a voice note on WhatsApp; 15 seconds later the order is in the ERP with the customer's personalised price list.
15 min→15 sec
Operations
Fifteen seconds of audio
A technician on site sends a voice note: job status updated, materials deducted from stock, PDF report generated.
manual→automatic
Administration
Receipt photo → expense recorded
A photo or PDF becomes amount, supplier, date, VAT, category — filed. Quotes and contracts generated in chat.
minutes→seconds
Marketing
New product → content ready
The AI reads the product sheet from the ERP and generates 5 photos, a 30-second video with voiceover, social posts, an SEO article and a newsletter.
2 weeks→10 min
Intelligence
Ask, and get real data in 8 seconds
“March revenue?” → €342,500 (+12.3%), top clients, margin, overdue invoices. Real queries on the ERP.
2–4 hrs→8 sec
Autonomous engine
You arrive — the company has already worked
Every morning at 7:00: emails processed, anomalies analysed, reports ready, the day's 3 priorities on WhatsApp.
07:00→already done
Swarm
Five AI executives every morning
A digital CFO, Sales Director, HR, Legal and CEO confer and produce a prioritised action plan. Automatically.
meeting→plan ready
Headcount analysis · The truth about departments
How many departments are overstaffed?
How many could you run with one person plus AI? Two cost lines, side by side.
| Department | Today | With ImpresaPilot | Savings/yr |
|---|---|---|---|
| Marketing3–5 people → 1 supervisor | €175,000 | €38,000 | €137,000 |
| Sales4–8 people → 2 performing like 8 | €280,000 | €72,000 | €208,000 |
| Annual total | €455,000 | €110,000 | €345,000 |
Note — If you don't have the department yet, even better: start with AI from day one. Three people instead of thirteen, and 10× the speed on content.
Digital workforce · The staff register
28+ digital employees. 14 departments.
Not generic chatbots. Every AI Worker has a role, a department, a dedicated prompt, a model, permitted tools and its own knowledge base.
Org chart — excerpt
- Digital CFOFinance
- Sales DirectorSales
- Marketing ManagerMarketing
- HR ManagerHuman Resources
- Legal CounselLegal
- Operations ManagerOperations
- IT ManagerIT
- Customer SupportCustomer Care
- Executive AssistantManagement
- AccountantAdministration
Roles in the product: ceo · cfo · coo · commercialista · finance · hr_director · legal · legal_counsel · executive_assistant · banking · bot_architect · bot_developer · bot_reviewer · bot_tester
Supporting documents · Exhibits
The product is real. In production.
The ImpresaPilot console, Admin Console v1.0. Real screenshots of the interface, with data from a pilot company in the food industry.
Income statement · The case on one page
The transformation, in figures.
Operating cost · traditional model
€450,000
per year
With ImpresaPilot
€48,000
per year
≈ €400,000 saved per year · estimated ROI above €400,000/yr
Comparative summary — annual cost
- Marketing — in-house department (3–5 people) €175,000
- Sales — in-house department (4–8 people) €280,000
- Cost of the traditional model €450,000
- AI platform + supervision (with ImpresaPilot) €48,000
- Annual savings ≈ €400,000 · −85%
| Operation | Before | After | Change |
|---|---|---|---|
| Order entry | 15 min | 15 sec | −98% |
| Revenue analysis | 2–4 hrs | 8 sec | −99% |
| 147 emails triaged | 1–2 hrs | 15 sec | −99% |
| Marketing content | 2 weeks | 10 min | −99% |
Price list · Terms
Transparent. As a prospectus should be.
Eras are one-off payments, financeable over up to 60 months. Subscriptions are billed annually. No hidden costs.
| Era | Duration | List price | Founder −50% | Early 100 −30% |
|---|---|---|---|---|
| Era 1 · Ignition | 2 months | €9,800 | €4,900 | €6,860 |
| Era 2 · Orbit | 4 months | €18,500 | €9,250 | €12,950 |
| Era 3 · ImpresaPilot | 6 months | €24,000 | €12,000 | €16,800 |
| Full Package | 12 months | €52,300 | €26,150 | €36,610 |
Standard
€490/month
Founder €245 · Early €343 · annual billing
- Updates, patches, new modules
- Email support within 24h
- 2 calls per month
- Proactive monitoring
Plus
€1,800/month
Founder €900 · Early €1,260 · annual billing
- Everything in Standard, plus:
- 4h/week of a dedicated developer
- Custom tools, tailor-made integrations
- Dedicated Slack/Teams channel · priority releases
Early Access · Founder
−50% on everything, locked in forever
First 10 · 7 seats left
Early Access · Early 100
−30% on everything, locked in forever
First 100 · 93 seats left
Subsidised finance · Year 2026
A 4.0 intangible asset. Stackable incentives.
ImpresaPilot qualifies as an Industry 4.0 intangible asset under Italian law. The incentives stack — and in some scenarios the net cost approaches zero.
| Hyper-depreciation 2026 | 180% |
|---|---|
| Nuova Sabatini 4.0 | grant |
| Digital Transformation MIMIT/Invitalia | 10% + 40% |
| Southern Italy Skills Development | up to 70% |
| Regional ERDF Vouchers | up to 100% |
| SIMEST Digital Transition | financing |
Net cost — scenarios
SME, Southern Italy
€48,000 → ≈ €5,944 net
SME, Central-Northern Italy
€48,000 → €0 net (ERDF voucher)
Estimated ROI ≈ €400,000/yr
Interface · Seven channels
You write to it like a colleague.
No software to install, no training. The AI lives inside the tools your company already uses.
- Telegram
- Slack
- Microsoft Teams
- Signal
- Web Chat
Security & governance · Technical note
Data never leaves your perimeter.
On-premise, GDPR by design, enterprise authentication. The system keeps working offline, on local models.
- True on-premise — Docker, 15+ services. No cloud SaaS.
- Authentication — OIDC · LDAP · SAML2 · Shibboleth · Active Directory · Azure AD · Okta.
- Role-based DLP — local models for sensitive data, never outside the perimeter.
- Full audit trail — every action by an AI Worker is logged and verifiable.
- Continuity — core functions stay up offline on local LLaMA/Qwen models.
- Liability insurance — coverage for data breaches and malfunctions.
Photographic appendix · The complete plates
The functions, one by one.
Before the prospectus closes: every console function, with its real screenshot and how it works in practice. Data from a pilot company in the food industry.
Plate I · Metric Intelligence
The operations dashboard.
EBITDA margin, days sales outstanding, inventory turnover, customer acquisition cost: your company's KPIs, computed on real ERP data and always up to date.
How it worksThe engine connects to your ERP (API, 500+ connectors), discovers the measurable metrics on its own and keeps watch. When a number drifts off course, it flags it — it doesn't wait for someone to open a report.
Plate II · AI Chat
You write to it like a colleague.
“Latest unpaid invoices”, “Top clients and most profitable products”, “Production cost analysis”: questions in plain language, answers with your company's real numbers.
How it worksPick the Worker (or write from WhatsApp, Telegram, email, Teams): the Worker queries the actual tools it is authorised to use — ERP, invoicing, warehouse — and answers within seconds with verifiable data, not opinions.
Plate III · Workers Catalogue
Digital employees, ready to hire.
Administration, production, sales, back office, development: every AI Worker comes with a role, a department, a seniority and a set of enabled tools.
How it worksSearch the catalogue by department or seniority and activate the Worker you need. It arrives pre-trained: a dedicated prompt, an assigned AI model, its own knowledge base and only the tools its role allows.
Plate IV · Tool Catalogue
70 tools, with the rules in writing.
From chart and document generation to connectors for the ERP, email and external databases: every capability of the engine is a registered tool.
How it worksEvery tool declares its connector, version and risk level. Workers only use the tools enabled for their role; the sensitive ones require human approval. No black boxes: everything is listed, versioned, revocable.
Plate V · Daily Briefs
“My Day”, already prepared.
Every morning, every role finds its own brief: what happened across the channels, suggested actions for the day, priorities highlighted.
How it worksOvernight, the engine gathers chats, emails, orders and metrics; at 7:00 am it distils a brief for each role — from the production operator to the CEO — and delivers it on WhatsApp too. You arrive, and the company has already been at work.
Plate VI · Suggestions
The AI proposes, you decide.
“Promotional campaign for the herb tortelli”, “Monitor ingredient costs”: concrete proposals, with a priority, a category and the role responsible.
How it worksBy analysing margins, costs and sales trends, the engine generates optimisation suggestions. You accept them, reject them or set them in motion: every decision is logged and sharpens the proposals that follow.
Plate VII · Communications Intelligence
Emails that read themselves.
52 emails in the period, 27 flagged, 271 alerts generated: every incoming message is read, classified by priority and turned into action.
How it worksThe engine reads the connected mailboxes, understands the intent (an order, a payment reminder, a complaint), assigns a priority and — where it has the rules — acts: an order arriving by email can start the ERP entry workflow on its own.
Plate VIII · Swarm
Many Workers, one goal.
One strategic question — “which product should we push?” — and several AI Workers sit down at the same table: administration, marketing, operations.
How it worksLaunch a session with a goal and the participants you want. The Workers exchange analyses built on real data (margins, cost per kg, order trends) and close with an operational summary: what to do, why, and with which numbers.
Plate IX · Governance
Users, roles, audit.
People sign in with their company credentials; every action by the Workers goes on the record. Governance isn't an appendix: it's built into the product.
How it worksUsers are imported from LDAP or sign in via SSO (Shibboleth, Active Directory, Okta); their role decides what they see and which tools they can call on. In parallel, the Activity Log records every bot action: who, what, when, at what cost.
Along the Three Eras journey you also gain: Meeting Intelligence (Meet/Zoom/Teams), content and document generation, Computer Use for ERPs without an API, Dev Pipeline, autonomous mode with guardrails and fine-tuned local models. The full scope is in the descriptive document — ask for it on a call.
The objections, addressed
Straight questions. Straight answers.
Does the AI replace employees?
No. It automates repetitive work and frees people for strategic activities — the ones that actually create value.
Is our data safe?
Yes. On-premise, it never leaves your perimeter. GDPR, role-based DLP, a full audit trail and insurance coverage.
Do I have to install software?
No. You use WhatsApp, Telegram or email. Zero training: you write to it like a colleague.
What if the internet goes down?
Core functions keep working on local AI models (LLaMA/Qwen), behind your perimeter.
Can I start with Era 1 only?
Yes. Each Era can be purchased on its own and delivers a return before funding the next.
How does it integrate with my ERP?
Via API, 500+ connectors, or Computer Use (Playwright) for systems that don't have an API.
Closing · Next step
Let's analyse your processes together.
We'll identify the first high-return automations for your business. A concrete conversation, no strings attached.