Custom software and AI for insurance brokers

Digitup Studio projects start at €1,000 and are delivered from 2 weeks. Above that minimum, the price depends on what the tool has to do, which channels it works in (web, WhatsApp, email), which systems it connects to (ERP, email, documents) and how much information it handles. A simple website assistant sits close to the minimum; an internal system with renewals, claims and a client portal costs more. The quote comes after we understand your brokerage. Depending on the project, payment can be by milestones, spread over 6 or 12 months or as a monthly fee, as we explain in how we work.

Digitup Studio delivers from 2 weeks for small, well-defined scopes. A first working version can usually be validated within weeks if the scope is clear; a system that connects to the ERP, email and insurer information usually needs more phases. We do not work with a standard number of weeks: we set the schedule after seeing how you work and deliver in stages, starting where the impact is greatest, so the team uses something real as soon as possible. If you need something ready in under 2 weeks, we are probably not the right option. We describe how we split a development into phases in custom software development.

Yes, and that is usually the starting point: we do not want to create another silo. We can connect or sync the ERP or policy management software, the CRM, email, Microsoft 365, Google Workspace, Drive, WhatsApp, telephony, e-signature and other tools with an API. With the ERP and with insurers’ systems, the integration depends on them offering an API or a reliable data export, and we check this at the start, before committing to anything. If a tool works well, it stays: we build the missing layer on top or migrate in stages. Many of these connections are automations between systems you already have.

With measures designed from the start according to how sensitive the information is, which in a brokerage can include health data in lines such as life or health insurance. We work with permissions per user and role, separation of information, action logging, protected credentials, backups and human approval for important operations. The AI provider and model are chosen with you based on privacy, cost and type of task, and a private search over your documentation (a RAG system) queries only authorised sources. The brokerage decides how it complies with the GDPR and its confidentiality obligations; we do not give legal advice and we build to the requirements you give us.

Work that consists of moving information automates well: identifying the client who writes by WhatsApp or email, finding their policy, classifying and assigning messages, extracting data from PDF policies, flagging expiry dates, requesting the documents for a claim and preparing draft replies. A WhatsApp AI assistant can handle the repetitive work and hand the conversation to the broker when judgement is needed. What we do not automate is the professional decision: recommending coverage or insurers, closing a renewal or offering another policy to a client. The AI can prepare the comparison or the draft, but the broker reviews and decides.

The two partners of Digitup Studio, a technical studio in Madrid founded in 2019, run it directly: Stas Nikitin, CEO, responsible for product and project management, and Anton Zverev, CTO, who decides the architecture and leads development. They are your point of contact from start to finish, without handing the project to a junior team after the first meeting. When needed, specialists from our network in UX/UI, frontend, backend, data or AI join, but the people in charge do not change. So whoever listens to how your renewals and claims work is also the person who decides how it is built.

Not yet: there is no brokerage case study published on our website and we prefer to say so plainly. What we do have is more than 50 projects developed since we started, several with the same logic a brokerage needs: backoffices with statuses and work assignment, chatbots connected to proprietary data, RAG over catalogues and documents, and integrations between systems. For example, for Tasaciones.com, an official valuations platform, we built from scratch the client app, the backoffice and the state machine for each order. Published projects are in case studies, and on the first call we review which ones are closest to your brokerage.

With a 30-minute call with Anton Zverev, the CTO, with no commitment. You do not need to arrive with a solution in mind: it is enough to tell us how clients come in, how you handle renewals and claims, what software you use and where information gets lost. We review the case, propose a technical approach and phases, and also tell you if we think it is not worth building. If it fits, we start with a focused first version where the impact is greatest, for example WhatsApp or renewals. You can book the call from this page or, if you prefer an earlier analysis of your processes, start with an AI audit.