Prototype to Launch

$5,000, fixed. If you move forward with the build, the entire amount credits toward your first month. The assessment then costs you nothing beyond the time it takes. If you decide not to build with us, you keep the plan and we part ways. There is no retainer and no minimum commitment.

Yes. We take AI-generated prototypes and rebuild them into healthcare products that can legally hold real data. We work with prototypes from Lovable, Bolt, Replit, Cursor, Claude Code, Codex, v0, Windsurf, and Base44. The tool matters less than what it generated. Every one of them leaves the same gap, with a strong interface sitting on a missing data layer, no audit trail, and no vendor agreements in place. MindSea has built software since 2007 with a North American team.

Usually not entirely. Your screens, flows, and product logic tend to survive into the build, because they represent decisions you’ve already made and tested on people. We almost always rebuild the data layer. Your AI tool built authentication, permissions, storage, logging, and hosting to make a demo work, and a security review will find them. The assessment tells you which is which before you spend anything on engineering.

Not as built. EHR integration runs through HL7 or FHIR, and AI coding tools don’t generate that architecture. The technical work is only part of it. Connecting to Epic, Cerner, or a regional system also means agreements, sandbox access, and review processes, and those run on their schedule rather than yours. We scope what your specific integrations require, including what each one adds to cost and timeline.

You need signed BAAs with every vendor touching PHI, encryption at rest and in transit, and role-based access controls. You also need audit logging of every record view and change, a documented breach notification process, and workforce training. The technical pieces are the fast part. The agreements and documentation are what take time.

Healthcare builds we scope typically land between $100,000 and $300,000. Integrations, compliance scope, and the number of user types drive where you fall in that range. An AI prototype usually lowers the cost of scoping, because you’ve already answered questions that normally take weeks of discovery. It rarely lowers engineering cost, because our team still builds the foundation to healthcare standards. The assessment gives you a range per phase that you can actually budget against.

You may hear “90 days” a lot. It usually turns into 4-6 months once security review, integration testing, and app store approval get added. The plan takes two weeks and gives you real ranges per phase. What stretches a timeline is usually EHR sandbox access, security reviews, and app store approval for regulated categories, and those run on someone else’s calendar.

It depends on whether your product carries regulatory weight. You’ll see offers to build an app in a month for a few thousand dollars. For a wellness or fitness product that never touches protected health information, connects to no EHR, and makes no clinical claims, that can be the right call. Once you handle PHI or integrate with clinical systems, the requirements change. You need a team that signs a BAA, carries insurance, and answers the phone when a health system sends a security questionnaire eighteen months from now. Rebuilding in-house works if you already employ engineers who have shipped regulated software.

It may qualify as Software as a Medical Device, which brings a different regulatory pathway and lifecycle requirements. This changes cost and timeline substantially, and founders often reach us without knowing it applies. We flag it in the assessment.