top of page


Our Story

Hi guys, folks have asked me to put a blog together to tell our story - answering key questions about who we are, where we're going, what we're trying to achieve, etc. 

Cliff Notes
After 10 years observing the challenges in getting small to midsize biotechs to adopt AI and with the new onrush of AI techniques which are being adopted aggressively by large pharma and AI-first biotechs, several of us decided that something had to be done - to fix what is effectively a broken pipe between bench scientists and data scientists.  This grass-roots effort is about impact and jobs-to-be-done; everything is open-source and the deliverables are targeted packages of {code + public data + education + workshops}.  We win when the smaller guys can and want to take advantage of AI/ML.  Everybody is welcome.




Now for the long version… 

Two sides of the table
The story starts about 10 years ago. I've been doing a variety of work, trying to get biologists in drug discovery to adopt AI techniques with some very reserved, limited success. Part of this issue is the (not-ready-for-prime-time) state of AI technologies, etc. And part of it is in actually looking at complex problems seen in early drug discovery, which is a kind of black box world when you get there, with lots of people doing things in a very subjective manner, based on how their lab operates, how their lab director operates, or how they've operated for the last 30 years. 

What I noticed is that there are two sides to the table.  One side is building the AI/ML etc to be useful. And on the other side, you have folks who don't want to touch this stuff, because they've never done it before. Between them you have inertia … 

The broken pipe problem
What happens is, there is a kind of fog that sits between them.  The bench scientists don't get the opportunity and they don't have the team members or the wherewithal to find out. And the vendors on the other side are overselling with the buyers confused and really being very skeptical.  Both sides are doing a great job trying - don't get me wrong. But both sides are missing each other at night. 

I have been observing this problem over the last 10 years and more recently at my current startup. I realized, we have to have a go at solving this problem.  There's a point where you see something and nobody's really paying attention, so you rally the troops. You go out and you talk to folks and you say, look, let's do something about this. There has been a lot of support to respond in a grassroots fashion to solve the problem by helping these two sides come together in what is a broken pipe problem.  You have a broken pipe between the two parties - smaller to mid-size wet lab focused biotechs and AI/ML vendors/providers. 

The broken pipe problem doesn't affect the larger companies, who have the wherewithal to hire in computational teams and create computational centers of competence. Additionally, there are a number of AI-first, AI-driven biotechs who are always in the press. We know about those, but the problem really is the rest of the guys - the 1000+ biotechs out there who don't really know about AI, who aren't big enough to just buy in skills and don't really trust the vendors who often claim that they can do anything. These biotechs are living day-to-day, hand-to-mouth, trying to create wet lab data so they can then raise more money.  They're on a time & cost curve that is very challenging. And in fact, if one could get AI/ML into their hands, in a way that works, that would really help them out. 

There is an opportunity to create a temporary pipe, which allows the two entities to operate together and to experience value.  Then they can double down and build a stronger pipe between themselves. But you need to create momentum, almost a snowball effect. They need to get to the point of saying: the temporary pipe has got us to where we need, now we're gonna go to some advanced algorithms etc.  So this broken pipe syndrome can be fixed by putting a temporary one in place which is open source, understandable, and even basic -  allowing biotechs to gain confidence to go off and use it more aggressively to solve for runway problems. 

Our mission
This is our mission. Even with all the news about AI, the small to medium size biotechs have not changed their use of algorithms (for the most) in the last 30 years. It's an impact mission. And everybody involved is very excited about this impact. We want data scientists and scientists to work closely together and we want to provide deliverables which help them create that temporary pipe. 

Enablement Packages to create temporary pipes - as part of a Unifying Data Science Framework
You need some software, of course, but software is not enough. There are some really good groups out there who produce open source software. You could also do some communications. There are great communication / ecosystem groups out there.  What we need to solve the pipe problem is a jobs-to-be-done approach. You have to think about a “package” which contains multiple pieces - software tutorials, education, public data, industry tips (pros and cons), and also hands-on workshops to deploy them. As with car-driving lessons you learn how to drive a car, here scientists can learn how to play with these packages. And this all needs to be done in an free non-profit manner, where people don't feel that this is a vendor-buyer relationship. 

With this in mind, we have created an Enablement Package concept. A package is there to help create a temporary pipe.  And we have to create the right packages - relevant and targeted. Otherwise, you have a bunch of efforts going on, which may not land anywhere and may not make a difference.  You create the pipes needed by biotechs.  So identifying the pain point, and then focusing the packages to the pain points is important. And then there's another layer, which is with packages I can stitch them together into a workflow to get a 10x benefit within what we are calling a Unifying Data Science Framework.

Grassroots energy
This is all based upon grassroots excitement and energy. People have interests in certain topics and it's their interest that you take advantage of.  That's how we operate. We don't push people into anything. We ask people: Where do you want to contribute? What's your science or technological interest, as a biologist, as a business leader, as a data scientist? And then we see working groups self form. It's very enjoyable. 
This will lead to a real change in the industry because it is where people enjoy themselves, where people see the value, they're learning, they're contributing, and they have a voice. Then they'll come to the table again and again to contribute and influence more, even though they're very busy in their professional and personal lives.  A good comparable would be the Raspberry Pi initiative.

Finally, a call to action
Anybody who finds this interesting, please sign up and participate in the effort. 

Thank you.


Raminderpal Singh

Take our 5 min survey

bottom of page