Season 4: Starting up

I'll be using this newsletter to document my journey of starting a company, at the intersection of AI and BI

This newsletter takes yet another pivot. There have been three “seasons” already, representing different parts of my career so far. And now that my career is going to take a new turn (yesterday was my last day at Delhivery), I’m going to pivot this newsletter as well.

Building in public

So I’ve been working on this idea (I’ll come to it in a bit, maybe in another post, but very soon) for a couple of months now. As part of this “work”, I’ve been meeting lots of people. Some of these meetings have been to get validation on the idea itself. Others have been with people who have built out successful B2B SaaS businesses, to understand how to go about finding customers, building and selling (I’ll talk about all of this in fair detail in subsequent posts).

More than one of these latter conversations (with people who have successfully built out B2B SaaS businesses) led to us discussing “content led marketing”. “You need to put out content”, people told me, adding that it can lead to inbound enquiries that are far more valuable than any outbound sales that I can do.

During one of these chats, I remarked that since I have been blogging for 19 years now, I might as well live blog this founding journey. The guy across the table, who I’ve known since 2005 but not really kept much in touch, overwhelmingly supported the idea, adding that “building in public” is a rather common thing to do for startups nowadays.

Over the following few days, I quickly got much more validation on this (“live blogging”), and so here I go, having waited out the period of my employment contract.

Why on this Substack?

I have started some seven or eight substacks till date. Most recently, I moved my flagship blog to Substack as well. So it would have been “natural” to start yet another Substack to document this journey.

However, thinking about it, what I’m planning to do is very much aligned with all the themes that I have covered on this newsletter. And given Substack’s cold start problem, I might as well repurpose an existing (and dormant) newsletter than start yet another new one.

So what is the idea?

As mentioned earlier, I’ll get to the specifics on the idea / product in subsequent posts, but for now I’ll say that this is at the intersection of artificial intelligence (AI) and business intelligence (BI). You might have forgotten since I haven’t written much here in the last 2-3 years, but one of my big lamentations about data science has been that the part that does “AI” (or machine learning, or any kind of learning) hardly talks to the part that does “BI” (getting insights from data).

So the hypothesis is that (as global as this sounds) using AI, we can do much better BI.

“Better BI using AI” was a tagline I quickly came up with. I had even decided to call my product “Minus AB” (Ai x Bi - get it? ), but it has got shot down from multiple sources, not the least by my wife, who says it “doesn’t sound positive” (obviously!).

So I’m still in search of a name.

Created using OpenAI Dall-E 3. I asked for an image to show “better BI using AI”

What to expect from this blog?

Given that I’m pivoting this blog again, it makes sense to set expectations clearly. There will be two parallel tracks that will keep running here - the process of starting the company, and the idea itself.

And in terms of the idea itself, given that it is at the intersection of AI and BI, and aimed at businesses, to be sold as a SaaS, you can broadly expect the same kind of themes that I’ve covered in this newsletter so far.

I’ll be talking a lot about dashboards, and “classic business intelligence”.
I’ll be talking a lot about data engineering, especially semantic layers and metric stores.
As you might have come to expect from me, I’ll talk a lot about communication and storytelling using data, and thus about visualisations.
We are in 2023 (soon to be 2024), so obviously I’ll talk about large language models (LLMs).
I’ll talk about data science and machine learning, and explainable machine learning.

And so on and so forth. It’s almost like I’m bringing everything I’ve done under one roof for this idea of mine. The one thing that potential investors can’t doubt about my idea is (I learnt this phrase recently) “founder market fit”, or should I say “founder product fit”?