EU funding
Funda
Reads EU funding calls as they publish and matches them to the organizations that fit.
- 4
- Role surfaces
- Admin, consultant, corporate, director
- 3
- Languages
- Every surface, not just the marketing pages
- 2
- Source pipelines
- SEDIA and TED, read by agents
What it does
Funding you qualify for, before the deadline
EU funding is public money that is hard to find. Calls land on SEDIA and TED as long documents with their own vocabulary, eligibility rules and deadlines. Most organizations hear about the right one too late.
Funda watches those sources on a schedule, turns each call into a structured record, and puts it in front of the organizations that match. Consultants work a portfolio of clients against the same feed. One dataset, read four ways.
- Admin
- The people who run the platform
- Consultant
- Advisors carrying several client organizations
- Corporate
- The organization applying for the funding
- Director
- Oversight across an account
How matching works
From a published call to the right inbox
- 1
Read the sources
Agents pull calls from SEDIA and TED as they publish, instead of waiting for someone to check a portal.
- 2
Turn documents into fields
Each call becomes a record: who may apply, for what, in which regions, by when.
- 3
Score against the organization
An organization's profile is compared field by field — sector, size, and where it actually operates, which is where the geospatial part earns its keep.
- 4
Route it to a person
Matches land on the surface that role uses, in that user's language.
What it shows
Same shape, smaller footprint
Voyager is where we built retrieval that stays grounded in a client's own content. On Polis, the hard part was an audit trail that makes a claim checkable. Funda carries both at a smaller scale: ingestion that has to be accurate, matching a user can argue with, and four roles reading one typed dataset.
The stack is deliberately small — Next.js on the front, Convex for data and functions, Clerk for identity and roles. The hard part is keeping an agent pipeline honest when the input is a call document written in institutional language and the output is a decision someone will act on.
Have a system with this shape?
Tell us what it has to read, who has to see the result, and what happens if it is wrong. We will tell you what it takes.