Introducing Gumnut: the platform for photo intelligence

Edit [2025-06-09] my wife points out that the logo definitely looks like an acorn and not a gumnut, so back to the drawing board on that part.

Introducing Gumnut: the platform for photo intelligence

In my post on Building the foundations of photo intelligence, I included several screenshots showing what I and others had built on the platform so far, but I didn’t include any links for anyone to try it out themselves.

That’s because at the time, the platform was hosted anonymously on a Render subdomain, and it had no authentication!  Why?  Because neither a name nor authentication were necessary to achieve the two goals I had for my research prototype: 1) Is the technology mature enough for all of these use cases? and 2) Do I have the knowledge and skills to implement it?

However, after I started getting some more inbound interest in the platform, I decided that it was time to start making it more official.

What’s in a name?

My first order of business was giving the platform a name.  I started by brainstorming a list of words that made me think of light, photos, understanding, and intelligence — words like “prism”, “lens”, “curate”, and so on — but unsurprisingly, domains associated with those words were already associated with other photo products, as well as heavily squatted.

With a great deal of enthusiastic support from my wife (she doesn’t care about AI, but she loves naming stuff), I ended up choosing the name Gumnut.  There are a lot of gum (Eucalyptus) trees near where we live, and our kids love picking up their seed capsules and bringing them home.  It’s short, easy to spell, memorable (I think), and importantly — the domain name was available!  And just like those gumnut seed capsules, I hope Gumnut helps people and businesses unlock the hidden potential in their photos and videos.

So I bought the domain name, and went through the brutally tedious process of migrating everything over to it.  Today, I’m excited to announce that:

• SDKs are now hosted on the gumnut-ai organization in GitHub: https://github.com/gumnut-ai/photos-sdk-typescript and https://github.com/gumnut-ai/photos-sdk-python

• API endpoints have moved to api.gumnut.ai, and generated OpenAPI documentation is now available at api.gumnut.ai/redoc 

• The platform dashboard has moved to app.gumnut.ai

Locking things down for security and privacy

Along with making the platform more public, I also realized that I needed to ensure that any photos and videos processed by the platform remained secure and private.  So I added authentication to both the platform dashboard, as well as all of the API endpoints.  User identity management, authentication, and authorization is now handled by Clerk.  On the backend, sessions are authenticated via cookies or bearer token authentication.  In addition, I added support for API keys for machine-authentication use cases.

What’s next?

While all of the API endpoints are back up and running again, now guarded by authentication, there’s still more migration work to do: the process of moving to the new domain and incorporating authentication broke some of my previous demos.  In particular, both the Immich API compatibility layer and the MCP server need to handle authentication before they’ll work again.

Beyond the work to finish authentication, I also want to build out some end-to-end demos of new use cases, rather than simple proof of concept demonstrations.  I’m particularly excited about some of these demos, as I think they’ll showcase how the platform unlocks possibilities that were previously considered impossible, incredibly time-consuming, or too expensive.  Stay tuned!

Finally, I’m planning to spend some more time on go-to-market (GTM) to help me understand the market opportunities for a photo intelligence platform like Gumnut, and to inform where I should focus my efforts in the future.

Getting involved

Building this all from scratch has been an exciting journey for me, but I love working with and talking to other people interested in the space.

• If you know someone with a background in B2B Sales / GTM and a strong interest in the photos space, I’d be interested in working with them.  I have a few folks helping me out with GTM right now, but I think I would benefit from partnering with someone more focused on this aspect of a business.

• If you know someone with a background in Applied AI as it relates to photos and videos, I’d love to chat with them.  While I believe my background and experience gives me the fairly unique ability to build a platform for Photo Intelligence from scratch, I’d love to learn more from people with deeper experience on the AI side.

• If you have personal or business use cases for Photo Intelligence, I’d love to hear about them, especially if I haven’t mentioned them in this newsletter in the past.

• If you’d like to try building something with the platform, I’d love to hear from you.  I’m not making anything self-serve right now, both because the platform isn’t ready for scaled consumption and because I want to understand everyone’s use cases better and work with them directly as needed.

In all of the cases above, drop me a comment below, or send me an email.

Special thanks

Thank you everyone for continuing to read my posts, follow along, and engage with me on this journey!  Also, a special thanks this week to Travis Britton, for providing some helpful guidance on finding good domain names.