Synthavo

AI platform for industrial spare-parts identification and ordering

Synthavo, Stuttgart, Germany.

Request

To redesign the user experience of an existing AI web product, rebuild its frontend, and define the post-MVP scope of the platform together with the client's leadership.

Result

A reworked web application for machine onboarding, parts catalogue management, and version tracking — delivered on top of the client's in-house AI engine and backend — plus a locked-in product vision for the next phase of the platform.

Technologies

React
TypeScript
Vite
Material UI
Vitest
React Testing Library
TUS protocol
PapaParse

Result

A reworked web application for machine onboarding, parts catalogue management, and version tracking — delivered on top of the client's in-house AI engine and backend — plus a locked-in product vision for the next phase of the platform.

Results

A reworked customer-facing product

The customer-facing side of the platform was rebuilt as a modern TypeScript and React frontend on top of the client's AI engine and backend, with the points of friction in the existing product resolved.
A locked-in post-MVP vision

Synthavo came out of the discovery phase with a structured post-MVP backlog, defined user roles, and end-to-end flows for both the web and mobile versions of the product — a clear plan to build against.

Challenge

Synthavo is an AI company in the industrial after-sales space, building a SaaS platform that lets machinery manufacturers and their customers identify and order spare parts from a single photo. The client had a working AI engine and an in-house backend team, but the customer-facing web product was lagging — flows built incrementally were confusing users, machine registration uploads ran for up to a day with no progress feedback, and version management left users unsure which configuration they were viewing.

The client engaged Cogniteq to address two gaps: the customer-facing product itself, and the strategic question of what to build next.

Solution

The Cogniteq team led the UX redesign of the core user flows and rebuilt the frontend in TypeScript and React with full automated test coverage, working on top of the client's existing AI engine and backend. The redesign was driven by real customer feedback from Synthavo's user base, covering machine onboarding, parts catalogue management, and version management. To keep frontend work independent of the client's backend timeline, the team built a contract-first API mock and specification.

In a later phase, Cogniteq's business analyst worked with the client's CTO to define the platform's next stage, producing a decomposed post-MVP backlog with the user roles, models, and end-to-end use cases needed to plan against it — across both the web and mobile versions of the product.