Blog · Published · Updated
An AI knowledge base for insurance broker operations
Short answer
An AI knowledge base in a broker's office gathers the documents for the current case: the policy, its references and the schedules for the right period, and it names the missing document. The broker then knows what to ask for before preparing an explanation for the client, and interpreting the policy stays the broker's work. At Syntalith the app costs from €3,000 net. At 500 questions a month of 12 minutes each, it pays back in 1.0 month.
How we workDetails ↓
- 01Call30 minutes with an engineer, free
- 02PrototypeA working system on your process and data, free
- 03ImplementationTarget and fixed price in the contract; 50% at signing
- 04Acceptance50% once the agreed target is met
- net per application, fixed price before contract
- from €3,000
- from signing the contract to acceptance
- 4–10 weeks
- free prototype on your documents
- €0
We belong to the OpenAI and Anthropic partner programmes and learn through their official courses. What this means ↓
Author: Syntalith · 3 min read
Statistics
Insurance and AI: statistics 2026
1.0 mo.
payback of a €3,000 app at 500 document questions a month: €3,490 of searching today, €534 with the app. Scenario.
15%
According to Brynjolfsson, Li and Raymond (2025), customer support agents working with AI resolved 15% more issues an hour on average, in a study of 5,172 people.
Source: Brynjolfsson, Li and Raymond, The Quarterly Journal of Economics (2025) ↗
47%
According to the European Commission (2026), customer service is the most common use of AI among EU SMEs and mid-caps that already use it: 47%.
Source: European Commission, Flash Eurobarometer 578 (2026) ↗
20.0%
According to Eurostat (2025), 20.0% of EU companies with at least 10 employees used AI.
Source: Eurostat, 20% of EU enterprises use AI technologies (2025) ↗
49%
According to Anthropic (2024), combining contextual embeddings with BM25 search cut failed retrievals by 49%, and by 67% with reranking.
Figures checked: 22 October 2026
The schedule found belongs to an earlier period
An employee at the broker's office is looking for a document about a client's new business location. The current policy refers to a schedule of locations, and the folder holds a file with a matching name, but from the previous period. The schedule the current policy refers to is missing.
The broker asks the person handling the case for a specific schedule, in place of asking in general for „the current documents” or searching the old schedule for the new site. A file missing from the searched collection does not settle whether the location is covered.
Help starts with assembling the right case documents
An employee taking over a client has to understand which documents make up the material for the question. A list of file names is not enough when each period leaves behind similar policies and schedules.
- 01QuestionThe employee asks about the client's location.
- 02PolicyThe assistant finds the current policy and its references.
- 03PeriodEach document shows the client, the period and where it was found.
- 04GapThe missing schedule is named, with a request to the person handling the case.
- 05FilingThe new schedule is filed with the right policy and period.
The schedule may still sit in correspondence that was never added to the library. The brokers decide which collections the assistant searches and who has access to them. The people who keep the documentation show where new schedules go and how the team finds them, because their work decides whether the next employee opens the schedule straight from the policy.
We run the trial on cases the office has already handled: the client's question, the set of documents and a note of what was missing. We compare the result with what the broker found by hand and measure the time from the question to the complete set.
What a knowledge base for insurance brokers costs and when it pays back
- Today
- 500 questions × 12 min = 100 h (€3,490)
- With the system, a person checks the output in 15% of the time
- €524
- Model, 1 call per case, Claude Sonnet 5.5
- €9.87
- Cost of one case
- €6.98 today, €1.07 with the system
- Saving
- €2,956 a month
Payback of €3,000
1.0 months
The volume, the time and the checking share are the scenario's assumptions. €34.90 an hour is the EU average hourly labour cost for 2025 according to Eurostat. Model: 8,000 input tokens and 600 output tokens per call at Anthropic's rates, converted at the ECB rate of 1.1205 USD per euro on 22 October 2026.
According to Brynjolfsson, Li and Raymond (2025), support agents working with AI resolved 15% more issues an hour on average. Less experienced agents resolved 30% more. In a broker's office the biggest difference shows for the person who takes over a client and does not know their history.
Use the document library you already have
The policy leads to all its schedules
a better case view in the current system
no new app
File version history is enough
versions in SharePoint
configuration
Documents in several libraries and in correspondence
an AI app that joins the sources
from €3,000
Microsoft describes version history in SharePoint, which lets you view and restore earlier versions of a file. The team still has to agree how a schedule is assigned to a policy and a period. When the problem appears later, while addresses are copied from a form, the post about reading data from insurance forms will help. The scope of the assistant is described on the private ChatGPT for your company page. For the call, bring a client question where the right document was missing last time.
Worked example · Scenario
How we count the payback: answers from company documents
We run the same count for your documents in the free analysis.
€3,723
a month for finding answers to 800 questions in documents
Payback: 1 mo.
Our assumptions are 8 minutes to find an answer in documents or ask a colleague, and checking in 15% of the old time. €34.90 an hour is the EU average hourly labour cost for 2025 according to Eurostat, for companies with 10 or more employees. Model cost: Claude Sonnet 5.5, about 6,000 + 500 tokens per answer from documents, at the ECB rate of 1.1205 USD per euro on 22 October 2026.
See the full calculation →
