Sundog Bio News | September 2025
Good morning,
September was an important month for us. We began it with a day-long workshop to agree both the product that we’re going to build for our very first customers - and how we’re going to work together to build it.
Alex Mitchell | CEO Update
Our vision is to make work better for biologists, and it’s worth remembering what the stakes are here. The 1.1 million life scientists working today are seeking fundamental insights into the biology of humans, plants and animals. The work those scientists are doing now, as you read this newsletter, has the potential to revolutionize healthcare - to save millions of lives - to feed billions. The people who are doing this work deserve software that works the way they do. That’s the goal we brought to the September workshop in which we defined what we’ll build for our very first customers.
We got clear direction from the biologists who saw our prototype. Everyone we spoke to collaborates with scientists in other labs, institutes and countries - and in other disciplines. Collaboration is joyful, as distributed teams share data and ideas that move the field forwards and get us all closer to solving the world’s biggest problems. As one of the scientists we spoke to put it, “we’re wired for collaboration.” In our workshop, we defined how the product we will build will be as well. We’re making it easier to share data, and easier to discuss it, wherever you are in the world. I’m confident in the network effects that will spread our product as it proves itself, again and again, in labs across the world. More exciting than that, though, are insights we’re going to make possible for the scientists in those labs. They will change the world.
Communicating those insights is almost as important as getting them. We also heard that that was harder than it should be. We want Sundog to make it easier to build scientific figures, whether that’s for high-stakes presentations like papers and posters, or frequent, informal communications like lab meetings. We spend time in our workshop defining what would solve the problems we heard about in our user conversations - and we defined a couple of stretch goals that will delight our very first customers.
It’s important to define what we’re going to build. How we’re going to build it - how we’re going to work, disagree, and make decisions together - is an equally important conversation.
We started our workshop with “premortem.” It’s a deceptively simple exercise; the goal is to answer the question “why did we fail.” It’s always a powerful conversation. It gets nagging doubts into the open, and naturally leads to creative thinking on how to mitigate the problems behind those doubts. Something that each of us brought up as a concern is that we might fall out as a team. The conversation we had in response about how we like to work, solve problems and disagree has already served as a guide to us as we navigate our early - and highly consequential - product and technical decisions.
We’re confident that we’re making the right product for scientists - and we’re confident in our ability to build it together. Here’s Tom to tell you more about how that’s taking shape.
Tom Armitage | CTO Update
The output of the September planning workshop wasn’t a to-do list, but a high-level backlog, with a prioritisation that came down to “essential for 1.0”, “nice to have”, and “down the line”. That’s about the level of detail we can deal with right now.
The backlog is large, but also very high level. Taking a story from it and working out what to do with it usually ends up spawning more story cards. We also don’t allow cards to move from our “Backlog” to our “Next Up” column without a reasonable pass at a spec written on them, that we’ve been able to discuss - another action that usually spawns more cards. None of this is a problem - if anything, I am glad we are making the path by walking - but it means that at these early stages, what feels like a straightforward job can often turn out to be a foundational one, and a “straightforward” task can turn into several dense, linked ones.
I’ve spent the past few weeks working on some really foundational stuff: authentication, the concept of “labs”, lab affiliations, and inviting users to labs. Nothing to do with high-resolution slippy images, nothing to do with realtime multiplayer. But all really foundational stuff that I’d like to get as good a first cut of as possible, so we don’t have to think about it - or radically change our minds - when we’re working on the imaging tools. This work has involved a few meetings, and whiteboard sessions, to make sure we are all happy with the decision we’ve made, that it maps to our users’ needs, and that we’re aware of the trade-offs we’ve chosen. This part of the Sundog ontology, that maps directly onto how scientists are working together in the real world, can’t be an afterthought. A lot of ground has been shifted in the past three weeks as a result, but we’ve ended up with something that we’re all very happy to build upon.
Whilst I’ve been writing this, Richard has been building out our AWS infrastructure, defining it in code, making it easy to manage, alter, and replicate. After a summer prototyping on somewhat simpler stacks, we’ve chosen to move to a full AWS stack for the project, and to do so at the start of our build. Sundog has particular constraints on where and how it can be built owing to the nature of the files we work with - specifically, that our compute needs to be in the same internal network as our file storage, or else bandwidth costs will eat us alive. I likened the decision to “buying the right size house to begin with”; it absolutely beats trying to move house and do a loft extension mid-project.
Finally, I’ve set a regular slot in the calendar for Friday Demos as a way to close out the week. Friday Demos is one of the rituals I’ve liked the most in my working life. In it, the whole company sits down together at the end of the week, and we show each other something we’ve done. That might be code; might be design; might be something else. The point is to show work to one another, and celebrate it.
Friday Demos isn’t a deadline. If you don’t have anything to show, that’s fine, and you shouldn’t be rushing to finish things for it. But this can be freeing in a different way: rather than showing a beautiful finished feature, we can show work in progress, the furthest we’ve got, to reveal where our heads have been all week, why we’ve been quiet. As companies grow, or as projects require attention such that you don’t know what everyone’s been up to all week, it’s a great way to bring everyone back onto the same page before the weekend.
It’s important that it’s not a crit environment - it’s not for evaluation, or gathering feedback. We don’t leap for things to improve, things to talk about next week, or ways we’d like to involve ourselves. But that doesn’t mean that improvements might not come out of it. We may stop ourselves from saying no but, but yes and is still acceptable.
And demos can be motivational for both the demo-er and the whole team. What you might think is unfinished, or janky, might be exciting to someone else - proof that something is possible, a demonstration of a delightful execution of an idea. It’s motivating to be reminded that yes, you have not finished this feature, but that doesn’t mean you’ve not been working hard and in a good direction.
Last week, I showed a first cut of “invitations” - inviting a user to a lab, letting them sign up to Sundog and then be added to that lab. And it worked, and it wasn’t there the week before. There were rough edges, but we don’t care about those - those will be filed out. At the same time, as I talked Alex and Richard through it, I felt that it was “all wrong” - I’d made some bad choices, prioritised the wrong paths. But we talked about those, and what we’d all learned seeing it in action, and planted the seeds for some course-correction the week after, and it turned out: it was a small course-correction, an important one, and we’d all learned something by watching the demo. Talking beats not talking, and showing work beats talking. I’m looking forward to future Friday afternoons as we share work with one another.
This month, we’ve done important work on both the product and the company; both what we’ll make, and the culture in which we’ll make it. We’re so glad you’re following along with us. As always, if you want to talk with us, reply to this email or get in touch with alexandra@sundog.bio.