I started Pakodas to talk about ML. But at times I feel strongly about certain ideas which are not ML. I will be writing short posts on them and you can choose to skip reading them. I will mention Nuggets in the title to show this is a general post.
Welcome to the new series => Nuggets => small posts of curiosity
You must be crazy if you decide to join a startup. You don’t even know what lays ahead. Maybe that’s why you join them.
I see young people joining startups without understanding how to evaluate them.
This is a simple heuristic to evaluate startups. It is designed for simplicity with high precision and not high recall.
Join a product based company filled with engineers from great product based companies where one of the founders has a tech background and the startup has more than a year of runway.
Why product based companies?
I have never worked in a service-based company but based on what I heard, it is just difficult to do great work in service based companies. Instead of trying to prove this, let me call it a personal bias.
Why engineers from great product based companies?
For startups to succeed, they require velocity.
To have high velocity, they should have a good stream of ideas.
To convert ideas to reality, they should have high deploy rate and low defect rate.
To have high deploy rate, you need to write less code via abstractions and have automations(CI/CD). Automation is what differentiates good from great.
To have low defect rate, you need to have experienced coders.
These skills are inherent in those who have worked for good product based companies.
Why one of the founder needs a tech background?
2 times I worked with founders who did not have a tech background. 2 times I worked with founders who had a tech background. I just don’t understand how can you make a tech company without a tech co-founder. IMO the quality to make good tech decisions of tech founders is way better than that of non-tech founders.
You can indeed make a great company like Cred and Zomato with non-tech founders but the choice is yours.
Why the startup should have more than a year of runway?
Joining a place with a crunch of cash means trying to save a drowning boat. Such companies have to optimise for survival instead of innovation which leads to short term thinking. It’s ok if the company is going to raise a new round soon which will give a year+ of runway. Avoid loss-making bootstrapped startups because the founders are under tremendous stress.
You can tick on all the above criteria and still end up at a shitty job if you get a weird job role. Understand, this is a filtering heuristic.
Come join Maxpool - A Data Science community to discuss real ML problems!