We found Urgency... or maybe, it found us?
The last few months something material has changed in our traction - we found the urgency every startup seeks. The spark that starts the fire. You can FEEL it.
A few months ago, we shared the success formula that drives our decision making:
Success = 5VU² - 2C - T
Where V is for Value, U is for Urgency, C for Complexity and T for time commitment.
As a startup, we keep navigating our product roadmap and our customer interactions to drive up the result of this formula:
We monitor how our customers use our platform, and engage our users constantly, to drive up Value.
We look for ways to reduce the Complexity of using the platform (for example, by integrating into their existing interfaces).
We drive down the Time commitment needed by the customer by doing more of the heavy lifting ourselves (via AI, or human labor).
Urgency, though, is much harder to control. Even though sales leaders always preach about creating urgency, that has very little impact. People have urgency to do certain things, and you can somewhat influence that (by telling them this offer will expire in 5 seconds), but the core of their urgency is intrinsic to their own priorities and plans.
So, as a startup, you try to find the urgency. You try to figure out what is top priority for your customers, and see if you map to that. You continue searching for it, and evolving your product roadmap, your pitch, your story, to find it.
Over the last couple of months, something flipped. Urgency found us.
I would like to thank the academy Google, Databricks and Snowflake
Back in early 2024, when we started our journey, it was obvious to us that AI will change data and analytics as we know it. That’s why we started in this space. How, and when, was still unknown. I think it’s still somewhat unknown today, but there’s a direction. A hunch. Something people are starting to realize.
With Google’s BigQuery for Gemini launch, and Databricks and Snowflake doubling down on their own (Databricks One and launch of Snowflake Intelligence), companies are starting to realize that leveraging AI for Analytics is possible. It’s hard. Sometimes messy and with weird results. But it’s starting to come within reach. We’re standing on a chair and extending ourselves to reach that amazing fruit on the tree. We’re not yet fully there, but we can see ourselves grabbing it one day.
So the big companies are creating a huge amount of hype around this, that’s great, but on its own, does not drive urgency.
What is helping drive urgency is that the business stakeholders, who are the D&A leaders’ customers, are also pushing strongly for this. Never before has the business pushed so strongly on D&A leaders to adopt a certain technology or use case. AI is different. Business users use ChatGPT regularly, and want to use it with their data. They hear their friends at other companies doing it, and they and they want to, too.
But, uhhh… there’s a problem. To get AI for Analytics to work, you need to teach it how your data works, your business context, what to use and what not to, and more.
Gemini for BigQuery requires that you document your data and profile it. Databricks recommends that you document your data and provide good queries to use. Snowflake asks you to fill out a YAML file, that should contain the documentation, JOIN information, sample data an verified queries. All of them make it evident one way or another that it only works well when you start small, and give it really, really, really, good documentation.
So wait… to leverage AI, humans need to explain all the data to it? But humans suck at documentation!
And that’s where an auto-generated semantic layer comes in
Data engineering teams now realize that there’s a massive pressure to deliver AI for Analytics, but to achieve it they need to work really hard. It will take them months, or even years, to do all the work the big platforms require of them to get this to work.
Or… they can get a high-quality, auto-generated, semantic layer. Over the last one year (plus) we invested most of our resources into building an auto-generated semantic layer that really works. It’s not perfect. It’s not bullet proof. But it works really well, it’s Solid. It takes good guidance from humans on how to be better. It will do the 90% of the effort, and work with your data engineering team to do the rest.
Solid can shorten the time it takes to get to true AI for Analytics considerably.
Between that, and the forces I mentioned in the previous section, we found Urgency. Or did it find us? Or maybe we both found each other like true soulmates?
Either way - we’re here and ready to support data teams on this journey. If you’re looking for an auto-generated semantic layer to accelerate your path into AI for Analytics - reach out.