sci2sci
A Berlin startup building a neurosymbolic memory layer for regulated enterprises.
Europe's most interesting startup stories. Every week.
No spam. Unsubscribe anytime.
Podcast Episode
🎙️ Deep Dive
Mistral AI, The Exploration Company, sci2sci — €3B for AI, $450M for space and verifiable systems — Daily News September 8, 2026
Listen to the episodeAbout sci2sci
sci2sci develops Integrity Cortex, a system that turns documents, data and AI outputs into a verifiable knowledge network: each claim must link to its source and every conclusion to its premises. VectorCat connects data distributed across cloud storage, network drives and laboratory systems without requiring a complete migration. The company starts with biopharma, where untraceable information may cause regulatory delays and failed audits, and plans to expand into other regulated sectors. In September 2026 it raised a €1.2 million pre-seed round.
The Story
Founded in Berlin in 2023 by Angelina Lesnikova and Valerii Kremnev.
How sci2sci works
Business Model
B2B software for data governance, knowledge verification and controlled AI use in regulated industries.
Revenue Model
Software licences, subscriptions and enterprise deployments; pricing and terms are not public.
Products & Services
Key Products
- Integrity Cortex
- VectorCat
- Parseltongue
Core Use Cases
- Verification of AI-supported claims and conclusions
- Dependency tracking across sources and decisions
- Cataloguing distributed enterprise data
- Preparation for audits and regulatory submissions
- Data governance for biopharma and banking
Market & Clients
Key Customers
Biopharma companies and other regulated organisations that need traceable data, documents and AI-supported conclusions.
Geographic Presence
How sci2sci competes
Competitive Advantages
- A combination of machine learning and symbolic verification
- Explicit links between each claim and its source
- Downstream impact tracking when data changes
- Connections to existing systems without complete migration
How sci2sci grows
Growth Strategy
Expand the engineering team, deepen biopharma deployments and take Integrity Cortex into other regulated sectors, including banking.
Distribution Model
Direct B2B sales through enterprise deployments and integrations with customer data systems.
Moat (Defensibility)
Neurosymbolic architecture, the open-source Parseltongue framework, enterprise integrations, regulated-data models and accumulated knowledge of source dependencies.
Regulatory Context
Biopharma use requires audit trails, electronic-record integrity, system validation and data protection; claimed compatibility with 21 CFR Part 11 does not amount to regulatory approval.
Key Risks
- Reliability claims still requiring independent validation
- Integration complexity across inconsistent enterprise data
- Long sales cycles in regulated industries
- Liability associated with incorrect conclusions
- Competition from better-capitalised data and knowledge-graph platforms
Strategic Insights
- Verification may become an infrastructure layer distinct from the generative model.
- Starting in biopharma focuses the product on workflows where the cost of error is high.
Funding & Investors
Total Funding
Funding Rounds
- 2026 - Pre-seed: €1.2M
Lead: Heliad, IBB Ventures
Key Investors
Founding Team
Founders
Key Metrics
Lessons from sci2sci
- Enterprise trust requires audit trails and deterministic checks rather than average benchmarks alone.
- Claims of absolute reliability need independent validation across production use cases.
Similar Startups
Almetra
Almetra mounts AI cameras above production lines to turn video into operational data - cycle times, output, utilisation - without IT integration, with customers including Bosch, Siemens Energy, ABB and Aumovio.
Read Analysistelli
A Berlin startup managing B2C interactions across voice, chat, SMS, WhatsApp and email.
Read Analysis