AI implementation partner
AI systems for complex insurance workflows.
Blankspace helps insurance and financial-services teams turn documents and operational processes into reliable AI-enabled systems—designed with the people who will use them, and built to run in production.
Use Cases
From a submission pack to an appetite decision
Experienced in large risks including Renewables, Oil & Gas, Construction and Marine insurance.
01 Intake
Emails and attachments
Schedules of values, loss history, slips, engineering reports, quotations, risk presentations.
02 Extract
Technical data, with its provenance
Every field carries a reference to the source text and a note on the reasoning applied.
Auditable back to origin
03 Correlate
Across the whole pack
Data is compared across documents to establish what is relevant in context.
04 Assess
Against underwriting appetite
Opportunities and risks surfaced for the underwriter to act on.
- Interface
- Underwriters review extractions where they work.
- Validation service
- Scores extractions so manual review is the exception.
- Integrations
- Into your systems, or external ones.
Why Us
We are not a traditional systems integrator
A flat, self-funded team that sits with your underwriters and operations people, and ships the software itself.
01
No distance between the business and the build
We capture and prioritise requirements with the people who do the work, and translate them into software quickly.
02
Service as a software
Client-specific requirements are what make a solution work, so we build for them rather than around them.
03
Self-funded, so the client is the only stakeholder
No external investors with competing demands. Meeting your expectations is how we measure success.
04
Lean and cost-effective
A flat team where product and engineering are the same function. No account layer to pay for.
Data & Security
We handle your data securely
We operate in full compliance with the GDPR.
We follow established cybersecurity practice to keep your data protected.
Your data is backed up and kept available.
Residency
Hosted in a region we agree with you in writing.
Model training
Client data is not used to train models.
Deployment
Runs in the cloud, with high availability and security.
Audit
Every extracted field traceable to its source.
How We Build
We build robust systems
Production systems, not proofs of concept: agents where they earn their place, and the instrumentation, testing and deployment around them.
01
Agentic first
We build primarily with AI agents, using modern architectures so each task is handled by the most effective combination of models, tools and deterministic methods.
02
Observability
Every AI input and output is recorded, together with system performance, so exceptions can be traced, diagnosed and improved.
03
Reliability
Regression testing and model evaluation are part of delivery, so AI outputs remain consistent as the system evolves.
04
Multi-language
The same system operates across the languages your business uses, rather than as a separate build for each market.
05
Multi-region
Applications run across multiple availability zones, with the regional customisation your operation requires.
Team
Meet the founding team
Blankspace has been created by experienced individuals with a broad background, from finance and audit to civil engineering, with one intersecting point: a passion and solid experience in delivering high-quality software directly to clients.
David Caputo
David brings 20+ years of combined business and technology experience—from auditing global organisations to engineering production-grade B2C, B2B and AI systems. He turns complex operational needs into reliable software.
LinkedIn →
Christina Tang
Christina brings 8+ years of Python engineering experience, including 6 years building AI-powered document-intelligence systems for insurance. She designs production-grade extraction pipelines, LLM infrastructure, and reliable backend platforms - turning complex documents and operational requirements into dependable software.
LinkedIn →
Marco Beghi
Marco brings 5+ years of experience with Python, including production work on RAG/LLM pipelines for insurance and banking clients, backed by 8 years in offshore energy risk and certification (Lloyd's Register, DNV). He brings hands-on experience designing LLM infrastructure and translating technical/compliance requirements into production.
LinkedIn →
Contact
Start a conversation
Tell us about the workflow, the documents involved, and the operational outcome you need. We will reply to work enquiries.