On-premise AI that reasons like a domain expert — with full traceability, epistemic humility, and zero data leakage. Built for healthcare, finance, and legal.
Executive Summary
PRISM (Platform for Reasoning, Intelligence & Specialized Modeling) is a self-hosted AI platform targeting €2.3 trillion in regulated industries that cannot use cloud-based AI solutions due to data privacy requirements. The platform enables professionals to create their own Professional AI Twin that runs entirely on their infrastructure.
The only solution where AI thinks like you + knows your data + stays completely private on your infrastructure.
€2.3 Trillion Left Behind
| Industry | Privacy Constraint | Regulatory Risk |
|---|---|---|
| 🏥 Healthcare | Patient data cannot leave hospital servers | HIPAA, GDPR violations = €20M+ fines |
| ⚖️ Legal | Attorney-client privilege prohibits uploading cases to ChatGPT or Claude | Professional liability |
| 💹 Finance | Trading strategies, client portfolios, M&A data forbidden from cloud AI | Regulatory sanctions |
"I'd love to use AI for diagnosis support, but I can't upload patient records to any external service. It's illegal."
— Dr. Elena M., Psychiatrist, MilanNot just AI. Your Professional AI Twin.
Thinks like you. Knows what you know. Always private.
Configure personality, expertise, tone, and thinking style. AI becomes an extension of YOUR professional mind.
Runs on your infrastructure. Data never leaves your control. 100% compliant with regulations.
Upload documents. AI learns YOUR methods, frameworks, and expertise. Persistent memory that improves over time.
| Feature | Description | Metric |
|---|---|---|
| Agent Wizard | 5-minute setup to define domain, expertise, and behavior | 5 min setup |
| Personality Engine | Configure tone, reasoning style, formality, citation preferences | ∞ configs |
| Knowledge Base | Upload and index documents for AI reference | 50K+ docs |
| Persistent Memory | Remembers every conversation, learns preferences | Forever |
| Web Search | Access real-time information when needed | Tavily API |
| Multi-Tenant | One platform, unlimited isolated workspaces | Complete isolation |
| Multi-Language | Support for global deployment | 100+ languages |
Source: Gartner, McKinsey AI Reports 2024
Four converging forces create an 18-24 month window of opportunity:
Open-source models (Llama, Qwen, Mistral) match GPT-4 quality
GDPR, AI Act (EU), HIPAA updates make compliance mandatory
€1,500/month today vs €10,000/month in 2023 (85% reduction)
Professionals tried ChatGPT — it's generic. They want specialized, private AI
18-24 months before big players enter the private AI space.
| Feature | ChatGPT | Claude | Enterprise AI | PRISM |
|---|---|---|---|---|
| Data stays on-premise | ✗ | ✗ | ✓ | ✓ |
| Custom knowledge base | ✗ | ✗ | ✓ | ✓ |
| Personality & reasoning config | ✗ | ✗ | ✗ | ✓ Unique |
| Epistemic humility | ✗ | ✗ | ✗ | ✓ Unique |
| Full traceability | ✗ | ✗ | ✗ | ✓ Unique |
| Setup in < 5 minutes | ✓ | ✓ | ✗ | ✓ |
| Price < €500/month | ✓ | ✓ | ✗ | ✓ |
| Unlimited usage | ✗ | ✗ | ✓ | ✓ |
Only solution where AI thinks like you + knows your data + stays private + tells you when it doesn't know
| Alternative | Limitation | Cost |
|---|---|---|
| ChatGPT Plus | No custom KB, data to OpenAI, generic | €20/mo |
| Claude Pro | No custom KB, data to Anthropic, generic | €20/mo |
| Specialized AI SaaS | Limited KB, cloud-dependent, vendor lock-in | €100-500/mo |
| Hire Research Assistant | Human limitations, availability, training | €2,000+/mo |
| Build Your Own | Requires technical expertise, maintenance | €10,000+ setup |
| PRISM Professional | Full KB, data sovereignty, domain expert, unlimited, memory | €199/mo |
| Plan | Price | Departments | Per-Dept Cost | Features |
|---|---|---|---|---|
| Team | €749/mo | Up to 5 | €150/dept | All Professional features, Admin dashboard |
| Business ★ | €1,499/mo | Up to 10 | €150/dept | All Expert features, SLA 99.5%, Priority support |
| Enterprise | €2,999/mo | Up to 20 | €150/dept | Everything unlimited, 24/7 support, SLA 99.9% |
All calculations based on 730 hours/month continuous operation. Infrastructure hosted on Scaleway (Paris, fr-par-1) for EU data sovereignty.
| Service | Specification | Monthly Cost |
|---|---|---|
| GPU Instance | L40S-1-48G, NVIDIA L40S, 48GB VRAM, 8 vCPU, 96GB RAM | €1,022 |
| CPU Instance | PRO2-S, 8 vCPU, 32GB RAM | €160 |
| PostgreSQL | DB-DEV-L, 4 vCPU, 8GB RAM, 100GB SSD | €55 |
| Block Storage | 200GB SSD | €20 |
| Object Storage | 500GB for knowledge bases | €7.50 |
| Flexible IPs | 2× static IPs | €6 |
| Infrastructure Subtotal | €1,270.50 | |
| Tavily API | Web search (estimated usage) | ~€40 |
| Domain & SSL | Professional presence | ~€5 |
| Monitoring/Alerts | Scaleway Cockpit + external | ~€20 |
| Buffer (10%) | Unexpected costs | ~€135 |
| Grand Total (Safe Planning) | ~€1,750 | |
| Level | GPU Instances | Concurrent Users | Monthly Cost | Cost/User |
|---|---|---|---|---|
| Launch | 1 | ~20 | €1,750 | €88 |
| Growth | 2 | ~40 | €2,772 | €69 |
| Scale | 3 | ~60 | €3,794 | €63 |
| Year 3 Target | 5 | ~100 | €5,838 | €58 |
Scaling is linear, not exponential. Cost per user decreases as capacity grows.
Achieved with 9 Professional users at €199/month = €1,791
OR 1 Business company (€1,499) + 2 Starter (€198) = €1,697
| Scenario | Customers | Revenue | Costs | Profit | Margin |
|---|---|---|---|---|---|
| Minimum Viable | 5 Professional | €995 | €1,750 | -€755 | Loss |
| Breakeven | 9 Professional | €1,791 | €1,750 | +€41 | 2% |
| Target | 12 Professional | €2,388 | €1,750 | +€638 | 27% |
| Comfortable | 15 Professional | €2,985 | €1,750 | +€1,235 | 41% |
| Optimal | 20 Professional | €3,980 | €1,750 | +€2,230 | 56% |
| Scenario | Customer Mix | Revenue | Profit | Margin |
|---|---|---|---|---|
| One Company | 1 Business | €1,499 | -€251 | Loss |
| Company + Extras | 1 Business + 3 Prof | €2,096 | +€346 | 17% |
| Ideal (1 GPU) | 1 Business + 5 Prof | €2,494 | +€744 | 30% |
| Scale (2 GPUs) | 2 Business + 5 Prof | €3,993 | +€1,221 | 31% |
| Optimal (2 GPUs) | 1 Ent + 1 Bus + 5 Prof | €5,493 | +€2,721 | 50% |
PRISM serves two distinct markets with the same infrastructure.
Target Customers:
Characteristics:
Example: Aerospace Company Structure
Characteristics:
Same infrastructure serves both markets. Individuals provide high margins; companies provide volume and stability. Together they maximize resource utilization and revenue.
| Phase | Milestone | Status |
|---|---|---|
| Phase 0 | Concept & Research | ✓ Complete |
| Phase 1 | Core Infrastructure & vLLM | ✓ Complete |
| Phase 2 | RAG System & Knowledge Base | ✓ Complete |
| Phase 3 | Memory System & Personality | ✓ Complete |
| Phase 4 | Multi-Tenant Architecture | ✓ Complete |
| Phase 5 | Frontend & User Experience | ✓ Complete |
| Phase 6 | Agent Wizard & Onboarding | ✓ Complete |
| Phase 7 | Document Intelligence System | 🔄 In Progress |
| Phase 8 | Advanced Analytics & Monitoring | 🎯 Q2 2026 |
| Phase 9 | Enterprise Features & Scale | 🎯 Q2 2026 |
Currently in exploratory conversations with professionals in healthcare and legal sectors.
Most pre-seed founders show a slide deck. We can demo a working platform.
| Timeline | Hire | Role |
|---|---|---|
| Month 1 | +1 | Senior Developer |
| Month 4 | +1 | Sales / Marketing |
| Month 8 | +1 | Customer Success |
| Month 18 | — | Team Size = 4 |
Accept the loss for learning
Lock-in: Early users keep this rate forever
Margin: 40-65%
Margin: 60%+
Beta and Early Access customers keep their discounted rate forever. This creates loyalty, powerful testimonials, word-of-mouth referrals, and case studies for marketing.
| Metric | Year 1 | Year 2 | Year 3 |
|---|---|---|---|
| Customers | 25 | 100 | 300 |
| MRR | €5K | €25K | €75K |
| ARR | €60K | €300K | €900K |
| Infrastructure Cost | €20K | €50K | €120K |
| Gross Margin | 67% | 83% | 87% |
| Shareholder | Pre-Funding | Post Pre-Seed | Post Series A (Est.) |
|---|---|---|---|
| Founder(s) | 100% | 85.7% | ~65% |
| Pre-Seed Investors | 0% | 14.3% | ~11% |
| Series A Investors | — | — | ~20% |
| Employee Pool | — | — | ~4% |
| Round | Timing | Raise | Pre-Money | Post-Money |
|---|---|---|---|---|
| Pre-Seed (Now) | Q1 2026 | €250K | €1.5M | €1.75M |
| Series A | Q3 2027 | €2-3M | €10-15M | €12-18M |
| Series B | Q1 2029 | TBD | €50-80M | TBD |
| Month | Milestone | Target |
|---|---|---|
| 3 | MVP Launch | 5 beta customers |
| 6 | Product-Market Fit | 15 paying customers |
| 9 | Breakeven | €3K MRR |
| 12 | First Company Client | €5K MRR |
| 15 | Growth Mode | €10K MRR |
| 18 | Series A Ready | €15K+ MRR |
18 months runway to profitability → €2-3M Series A at 10-15x valuation
At Series A (€15-20M valuation), €250K becomes €2.5-3.5M
Entry at pre-seed valuation before market validation
Platform supports 100+ languages, multi-region ready
GDPR, AI Act, HIPAA all push customers toward privacy solutions
| Milestone | Timing | Target |
|---|---|---|
| Series A | 18 months | €10-15M valuation → 10x return |
| Series B | 36 months | €50-80M valuation → 40-50x return |
| ARR Target | Year 5 | €50M |
| AI Twins Created | Year 5 | 10,000+ |
| Geographic Reach | Year 5 | 50+ countries |
| Exit Target | Year 7 | IPO or Strategic Acquisition |
"In 5 years, every professional will have a Professional AI Twin that thinks like them, knows what they know, and never forgets.
PRISM will be the platform that makes this possible — privately, securely, worldwide."
"We're not building another AI tool. We're giving every professional a digital version of their best thinking — private, powerful, and always available."
Only privacy + personality + affordability in market
Built the platform, not just a business plan
GDPR, AI Act push customers toward privacy solutions
100+ languages, multi-region deployment
€250K runway to profitability
AI SaaS M&A at 6-10x ARR
before big players catch up in private AI space
AiDome Technologies OÜ
Tallinn, Estonia
Schedule a ConversationThis document contains forward-looking statements and projections based on current expectations. Actual results may differ materially. This is not an offer to sell securities. Investment involves risk, including potential loss of principal.