Latest Notes

Roman Grossi β€’ Founder

Indie hacking, startups, resilient systems - and staying sane while building a small company

Back to articles

Saylify Update: Fighting Perfectionism, Refactoring, and Finding the Right Focus

Β· 2 min read Β· 12 views

I have not written anything about Saylify for a long time, even though I planned to launch in January. Unfortunately, life likes to throw in challenges you can drown in.

But that is not the only issue. Here is what else I have been dealing with:

1. An inner perfectionist that keeps saying: 'You can do better! Add websockets here and get rid of long-polling!' Did I manage to tame it? Of course not. From time to time I give in a little so I do not burn out because of it.

2. Positioning problems. An engineer's brain is used to complex systems and loves when a system has a lot of toggles, parameters and secondary features. Sadly, that shows up when I build my own product too. I started making Saylify by trying to solve every pain at once without clear positioning. After a few sleepless nights of thinking and market research, I found an elegant solution: multiple entry points, one artifact. I sort of solve several user pains, but at the same time I do not overcomplicate the system. Instead I build an effective funnel that leads everything to one result.

3. Code architecture. In programming there is the DRY rule: 'Don't Repeat Yourself'. The first thing I did was break this rule and, without noticing it in the moment, I created 3 (THREE!) almost identical API endpoints. So I spent some time on refactoring, which by the way is still ongoing.

4. Lack of focus. The problems I am solving in Saylify are very interesting and closely tied to various AI things. Sometimes I dive so deep that I start researching secondary topics around LLM, ASR and even Speech Understanding.

So what about the update itself: I will keep the part I am working on now a little secret. Here is what I can say: the core of the project is 99% ready, and right now I am working on the client side that users will interact with. There is just a tiny bit left to do. At least I hope so.

Bottom line: I do not want to publicly announce any deadlines. I just keep building the project while balancing life challenges where failure could be catastrophic.

More to explore

Human-Like Memory for LLMs

TL;DR I wrote a manifesto-style essay about a memory model for LLMs that is as close as possible to human memory and lets the system build a relationship histor…

Jan 19, 2026 Read more