AI startups for the rest of us?
Finding the path to becoming an indie machine learning hacker
If you are someone who can write a line of code or two, then you can start a SaaS company.
You can rely on websites/podcasts like Indiehackers and Startups for the Rest of Us for guidance, thought-leadership, community support, and even finding partners. You can plug into a broad ecosystem of services to handle things like automation, customer support, and marketing. What starts as a side-hustle can turn, often within the space of months, into something that eclipses salaried income.
Some are content to build single person or small-team lifestyle businesses. Others seek to build a team and grow a software business that dominates a niche market. VC funding is not a fit and perhaps even poisonous. However, there is an increasing amount of external funding options to bootstrapped SaaS companies with small total addressable markets, where "small" means a late-stage valuation on the order of 10 to 100s of millions of dollars.
Is indie-SaaS for AI a thing?
But I have some questions. What is the machine learning/data science version of this model? More specifically;
What does the blending of data science/machine learning and SaaS look like?
How talented a data scientist/ML engineer do you have to be to make it work?
What are the models with unit economics that have a fighting chance of covering the opportunity cost of high compensation jobs from big tech companies?
Does all the good data live in silos?
Are the labor and cloud costs of labeling and training models too high for a bootstrapped company?
Who are the role models in building this kind of firm?
I've spent a few months interviewing various experts on these questions and will begin to publish my results in this newsletter. If you have any feedback, and if you know anyone who would be worth interviewing, I'd love to hear from you. Please reach out.
Some of the more nuanced and in-depth analysis I'll publish only to paying subscribers. More of the handwavy stuff will go in the free version.
If interested, sign up for a paid subscription within the next 24 hours for 20% off for the year. Or don't reach out, and I'll give you trial access.