Kapability AI reads adjuster reports, medical files and policy wordings, checks them against your guidelines and prepares each case for a decision. Your claims handlers and underwriters make the call.
4. Findings
The vehicle was inspected at the contracted workshop on 16 September 2026. Impact damage was found on the front right corner, affecting the bumper, headlamp unit and hood. The windshield is cracked across the passenger side.
The third-party driver accepted full liability at the scene. A signed accident report is attached as Appendix B.
The workshop estimate totals €8,420.00 including labor, of which €1,810.00 relates to glazing.
5. Recommendation
Repair is economical. No signs of earlier damage were found on the affected panels, and the parts listed match the impact pattern described by the insured.
The workshop has confirmed availability of the headlamp unit. Expected repair time is six working days.
Foundation models fine-tuned on insurance and banking data, then on yours. They know a deductible from a sub-limit.
Storage, processing and training stay on your own servers. Nothing is sent to third-party AI services.
Kapability runs as a layer above your core and document systems. Nothing is rewritten or migrated.
Payouts, policy approvals and risk acceptance stay with your people. The model prepares; your team decides.
Every claim and every application arrives as a stack of documents: adjuster reports, physician notes, repair estimates, policy wordings. Before anyone can decide, someone has to read all of it, page by page.
Handlers re-key fields from adjuster reports and check every line against the policy by hand. Payouts slow down, leakage creeps in and SLAs slip.
Underwriters comb through lab results and discharge summaries to find the few risk factors that matter, and applications queue behind them.
Replacing a legacy core takes years and puts operations at risk. Automating document work shouldn’t have to wait that long.
Each module is fine-tuned for one department’s documents and rules, and goes live on its own. Start with the desk where the backlog is longest, then add the next.
Motor, health and property claims
Kapability reads loss adjuster reports, accident forms, repair estimates and medical bills, then checks each finding against the active policy before a handler opens the file.
| Line | Amount | Policy check |
|---|---|---|
| Front bumper, replace | €1,240.00 | Covered |
| Headlamp unit, right | €960.00 | Covered |
| Hood, repair and paint | €1,690.00 | Covered |
| Windshield, replace | €1,810.00 | €310 over sub-limit |
| Labor, 14 hours | €1,260.00 | Covered |
| Paint materials and sundries | €1,460.00 | Covered |
| Total claimed | €8,420.00 |
Recourse possible. The third party admitted liability (adjuster report, p. 3).
Life and health applications
Physician notes, lab results and discharge summaries are parsed for chronic conditions, history and clinical indicators. Each finding is mapped to your underwriting guidelines and risk register.
| Finding | Source | Guideline |
|---|---|---|
| Type 2 diabetesDiagnosed 2019, HbA1c 7.1% | Lab report, p. 4 | M-12, refer |
| HypertensionControlled, 132/84 mmHg | Physician note, p. 2 | C-04, standard |
| Non-smokerCotinine test negative | Lab report, p. 6 | Confirmed |
| BMI 29.4178 cm, 93 kg | Medical exam, p. 1 | B-02, within table |
Decision: underwriter. Suggested next step is a referral under guideline M-12.
Portfolio, renewals and field teams
Customer profiles and portfolio data are analyzed for lapse risk and cross-sell potential, and field managers get a portfolio brief before every agency visit.
| Policy | Renews | Signal |
|---|---|---|
| Fleet, 42 vehiclesMotor | 12 Dec | Lapse risk high |
| Home and contentsProperty | 19 Dec | Offer earthquake cover |
| SME packageProperty | 3 Jan | Broker changed |
| Family healthHealth | 8 Jan | Dependants turning 18 |
Visit brief ready for 14 Nov. Four renewals to discuss and two cross-sell openings.
Service and support
Incoming emails, portal messages and chats are classified by intent and sentiment, routed automatically and answered by assistants grounded in your own policy documents.
“It’s been two weeks and nobody has called me about my claim.”
“Can I add my daughter to my health policy from next month?”
“Thank you, the payment arrived this morning.”
Drafted replies are reviewed. An agent approves every answer before it is sent.
One motor claim, followed from the moment it arrives. Medical files, policy changes and customer messages go through the same four steps.
Kapability sits above your policy administration, claims and document systems and picks up new files as they arrive. Nothing in the core is rewritten.
Models tuned on insurance data read scanned and digital documents the way an experienced handler would: fields, line items, clinical findings and liability notes. Every value keeps a link to its page and paragraph.
Findings are matched against the policy wording and your guidelines. Limits, exclusions, deductibles and recourse potential are worked out before anyone signs.
The handler opens a prepared case, traces any value back to its source in one click and makes the decision. Every step is logged for audit.
Kapability is deployed inside your network. Policyholder data is stored, processed and used for fine-tuning only on infrastructure you control, under your own security policies.
Zero calls to external AI services. The models run where your data already lives.
Runs on your own servers, inside your perimeter. Your data never trains anyone else’s model.
Connects above existing ledgers and databases, without core code changes or data migration.
Every extracted value points to the page and paragraph it came from.
Payouts, policy approvals and risk acceptance always need a person’s sign-off.
Designed for GDPR requirements. Access control and data retention follow your own policies.
A claims or underwriting project touches operations, IT and risk. Here is what each of them gets.
Backlogs clear and review cycles shrink from days to minutes, without adding headcount.
Underwriters spend their time evaluating risk instead of hunting through PDFs.
Document intelligence in weeks, as a layer above the core you already run.
Private execution, a traceable record and a person’s approval on every decision.
We work with one department at a time, alongside the people who do the work. When a rules engine or a simple script will do a job better than a language model, we say so.
We sit with the team, whether claims, underwriting, sales or service, and measure where the hours go.
We test whether a language model is the right tool for each task, or whether a simpler method wins.
Foundation models are tuned on our insurance and banking datasets, then on your data, inside your environment.
Each module is validated against your documents, formats and guidelines before it reaches production.
Modules go live one at a time on your servers. The next starts once the first has proven its value.
Something else on your mind? Ask us directly.
No. Kapability is deployed on your own servers. Storage, processing and fine-tuning all happen there, and documents are never sent to third-party AI APIs.
No. Kapability runs as a layer above your existing databases and core ledgers. It reads from them through the access you grant, and nothing in the core is rewritten.
We start from proven foundation models and fine-tune them on our own insurance and banking datasets, then on your data inside your environment. The model choice is part of the feasibility assessment for each module.
Your people. Kapability extracts, checks and flags. Payouts, policy approvals and risk acceptance always need a claims handler’s or underwriter’s sign-off.
We run Kapability on a sample of your real documents, such as loss reports or clinical files, and evaluate the results against your own guidelines. The PoC doesn’t connect to production systems and is covered by an NDA.
Yes. The same approach applies to banking operations, where documents are just as dense and the rules just as strict.
Request a technical proof of concept. We run Kapability on sample loss reports, medical files or policy documents in a private environment and measure the results against your guidelines.
Thank you, there. We’ll write to you to agree the scope and the sample documents for your proof of concept.