Greptile logo
How to Ship Faster, from the founders of Greptile (YC W24)

How to Ship Faster, from the founders of Greptile (YC W24)

February 7, 2023 (1y ago)

Daksh Gupta

We designed, built, and shipped the first version of Onboard at the Scale AI Hackathon in San Francisco on July 15, 2023 — in one day.

Soohoon and Vaishant at the Scale AI Hackathon, where we wrote v0 of Onboard AI Soohoon and Vaishant at the Scale AI Hackathon, where we wrote v0 of Onboard AI

3 days later we released it to the public.

8 days later we added Stripe and got our first paying user.

Since then, we have pushed new code every single day, and serve thousands of developers every week. There are certainly faster teams out there, but in my experience, we are certainly well above average in how quickly we ship. In fact, I truly believe our speed was easily the strongest part of our YC application.

Tweet from Sohan, founder of Flint (YC S23) Tweet from Sohan, founder of Flint (YC S23)

I think there are five reasons for this, and I have listed these below. If you’re a startup that wants to ship faster, this should be helpful.

1. Alignment-backed motivation

The greatest factor in productivity for ambitious people is how genuinely they believe that what they are doing today matters. There doesn’t need to be a lofty, change-the-world mission. Ours is to have millions of daily active users who love us. Every task we add to the list is there because there is a logically sound chain that connects it to that goal. This means whoever is assigned that task:

a) understands and appreciates our goal.

b) genuinely believes that their task is going to impact our trajectory towards it.

There probably shouldn’t be many tasks on the list that don’t impact our goal anyway, so this is also a good filter to prevent bogus tasks from clouding our time. Ambitious people don’t want to do things that they don’t think matter.

2. Beautiful products

Dirty homes get dirtier, clean homes get cleaner, and also get flowers and scented candles. A beautiful product inspires you to build more features and fix bugs. Beautiful products are beautiful not just for the end user but also for the developer. That means clean code that’s readable and inspires improvement. It should be a clean home that compels you to get nice furniture, change the curtains, and in turn keep it clean.

Beautiful products are easier to feel proud of and believe in. This is a surprisingly large leverage on productivity. Conventional wisdom is to not sweat the small stuff early on, but I would argue it takes only 10% more effort to make a beautiful product than a basic one, and the ROI is immediate.

Beautiful products are both the cause and effect of a fast, well-functioning team. If your product is beautiful, you will care about it more, and if you care about it more, your product will be more beautiful.

My intuition is that more of Stripe’s success than one would think is downstream of the fact that people like beautiful things.

Because what does a beautiful thing tell you?

Well, it tells you the person who made it really cared.

— Patrick Collison, co-founder of Stripe.

3. Clock speed

Our sprints are 1 day. Since we are all roommates, we do an Asana update usually on the bus to work together, and a retrospect on the bus back. This means our clock speed is at least 30 cycles a month, usually more as we will often retrospect and plan over lunch.

Shorter sprints mean more opportunities to evaluate that every task is truly necessary or at least valuable (see #1). It also means the time spent blocked is minimal. Lastly, it means we can dynamically reassign tasks if some task takes more or less time than anticipated.

4. Observability

We use Asana, the least complicated tool we possibly can that still lets us create, assign, schedule, and track the status of tasks. The process is simple — if we agree something needs to be done, we add it to the “TODO” column, assign a person, and set an aggressive but realistic due date. We only have two other columns: “Doing” and “Done”. Chances are, if you’re a small team like us, you don’t need much more.

When there is always an available pipeline of open tasks, no work time is ever spent inactive. Inactive time at work is poison. It is not a break, so it doesn’t replenish your energy; you are burning energy with no productive output. Energy and motivation are limited (but growable) resources. Their expenditure should be mindful.

5. Energy invested, not spent

Energy is a finite resource, and for that reason, most people view it as something that is either spent (by working and producing output) or conserved (by resting). I propose that there is a third option — energy can be invested and grown. There are certain productive tasks that create energy. These are different for different people. For Soohoon, refactoring is an energy-investing activity. Akin to cleaning one’s work surface or closing tabs and unused windows, it is a decluttering activity that quickly pays for itself in added productivity. Partly because it makes all new work more satisfying, and partly because it makes the product more beautiful and inspiring to work on.

There are other forms of energy investment too. For me, a quick sync with my co-founders, talking through our current problems and paths to success, even if just for a few minutes, yields mental clarity and quickly pays for itself in new productivity.

I think it’s worth your time to genuinely examine what an energy-investing activity looks like for you. What work creates net positive energy for you?


I hope this helps. I want to write on this exact topic again in 1–2 years and later when we have a larger team. I suspect I will adapt many of my opinions here as time goes on.

If you’re interested in what we’re doing at Onboard, check us out at

You can use code

for a free month of Onboard Pro or Onboard for Business.

Onboard helps you ship faster by accelerating debugging, understanding complex code, writing tests, and more. It is ChatGPT with a deep, nuanced understanding of your repos.

Ready to get started?

Sign up for free and start searching.