Software Engineer at J.P. Morgan. I can build things, but getting users? No clue.
Hey everyone.
I'm Jack. I work as a Software Engineer at J.P. Morgan Chase, mostly in portfolio optimisation and real-time financial data.
I've been thinking about building something on the side. The area that interests me is AI tooling for engineers. At work, I keep noticing people try out LLM-enabled development, hit a few early frustrations, and just... stop. They don't push through. I'm not sure if that's a product opportunity or just normal friction, but it's stuck in my head.
Honestly, I'm pretty early in my journey. No product. No audience. I've never really put myself out there before, and the whole "marketing" side of things feels alien to me.
So I figured I'd start here and see what I can learn. If you started from nothing, how did you actually find your first users? Any resources, people to follow, or tactics that worked for you? Also happy to just have a conversation if you've been through this before.
Happy to help if anyone has questions about Python, data pipelines, or working with LLMs. That part I can do.
Successful products and services do one of two things for actual people: reduce a pain or increase a gain. The people who will benefit are your target audience.
Figure out how many of them there are that you could ever possibly reach. Then how many of those you can actually reach. Then how many of those you might be able to get to convert into revenue.
An example for a mobile app for music :
Study your competition. If there is no competition, that's a red flag because it's unlikely that your idea is genuinely both unique and good.
Find them where they are. You'll meet more music app people at a concert or festival than at a tech conference.
Thank you for the advice! I appreciate it. I will start understanding my competition and target audience then.
Hey Jack, this is a great starting point! The best way to find your first users is to start with people who share your pain—other engineers experimenting with LLMs. Share small experiments, workflows, or learnings on X/Twitter, Reddit, or Discord. Document your journey publicly and engage with anyone who responds. Early feedback from a handful of real users is way more valuable than chasing a big launch.
Thank you Andy!
Jack, this is way more common than it looks.
Quick pattern I’ve seen with strong engineers:
You don’t fail at building.
You stall at getting something into users’ hands, then assume the idea’s bad.
What you’re describing around LLM tooling at JPM is interesting not because “AI”, but because friction + abandonment usually hides:
• unclear ROI
• bad onboarding defaults
• or small usability cuts engineers never complain about — they just drop
Early on, I honestly wouldn’t obsess over “marketing”.
The fastest signal usually comes from:
build something tiny → put it in front of 5–10 real engineers → watch where they stop.
That is distribution.
I work with a few technical founders who move faster by pairing solid execution with some GTM structure early.
If you ever want to sanity-check the idea or the go-to-market side, happy to chat.
Either way, you’re asking the right questions way earlier than most.
Thank you Ege for the insight! I appreciate the offer. That aligns with my experience. You said it much better. What I seem to be seeing is that the ROI does not seem clear and they are experiencing usability issues.
Hi Jack,
Maybe you can be one of Uclusion's first users. Engineers are very demanding customers and what you see with LLMs is an example of that.
Hi David, you are right about engineers being demanding users. I'm not in the market for new tools myself, but good luck with Uclusion.