Insights
Running claims with AI, told from the operation.
What we have learned taking AI into live claims operations: what to automate, what to measure and what to ask before you start.
Travel: a hospital bill in another language and currency, start to finish
From the first email to the recommendation: reading, conversion, coding, pricing and checks on one claim.
Pet: a vet invoice checked line by line against the vet's report
Fifteen lab tests on the invoice, six in the report. How each line is matched to the fee schedule and to the evidence.
Home: a water-damage claim where the photo does not show the wall
The invoice bills drywall and the only photo shows a wet floor. How photos are read as evidence, line by line.
Auto: a roadside assistance reimbursement, start to finish
Invoice, card receipt and assistance record, checked line by line on a sample case.
Health: a scanned urgent-care bill with no medical codes
Codes assigned, each line priced against the benchmark and a misread total caught by one check.
Case studies: Claims47 and our teams in production
Travel assistance, pet, health, home and more: what was deployed and what was measured.
Published in the media
What we publish
Opinion pieces and articles by the M47 team in insurance and technology media. Original language preserved.
Cómo llevar la IA a producción y mantenerla bajo control
AI does not improve just by connecting it to your data: it improves when you measure it, understand its errors and correct them. Five steps to take AI to production and keep it under control. By Marçal Albert, Principal Software Engineer.
Read the article →INESE · 2026-09Cuando el clima empieza a formar parte de la póliza
Heat waves, pollution and pollen already shape health and travel claims. Why environmental variables belong in the risk model and in claims analysis. By Jordi Guitart, VP of AI Health & Pharma.
Read the article →SegurosNews · 2026-07Cuando la póliza entiende el viaje
The next generation of travel assistance will accompany the traveler throughout the whole trip, not react after the claim. By Ángel Pavón, Head of Revenue and Strategy.
Read the article →El siniestro como problema de datos no estructurados
Read the article →Füture · INESELa gestión de siniestros en tres dimensiones: eficiencia, incentivos y balance en la era de la IA
Read the article →Füture · INESEDe la codificación clínica al insight actuarial: cómo la IA transforma la gestión del riesgo y la validación de claims en seguros de salud
Read the article →Füture · INESEDel bluff a la evidencia (con cambios)
Read the article →Füture · INESECloud sin gobierno, IA sin evaluación. Dos deudas que el sector asegurador acumula en silencio
Read the article →Interviews & features
Entrevista a Enric Plana, CEO: "El foco no es 'innovar', es operar mejor"
Read the article →Marketing4eCommerceEntrevista a Enric Plana, CEO de M47 AI Company
Read the article →Digital Biz MagazineEl impacto de la IA en el sector seguros
Read the article →Revista ByteIA en atención al cliente: adaptabilidad, ética y eficiencia
Read the article →Fundació Impulsa"El que abans feien amb cinc setmanes ara amb la IA es fa amb cinc minuts"
Read the article →A Call for Principle-Based Acceptance of Synthetic Patients: Establishing the AI-Driven Foundation for Regulatory Submissions
Read the article →Our column on Füture by INESE
M47 writes regularly on AI and insurance operations in Füture, the innovation outlet of the Spanish insurance sector.
All our Füture pieces
Send us a handful of your claims and we will show you what Claims47 does with them.
A 20-minute demo on synthetic data, or a test on your own anonymized files.