When I started building Rendi, I looked at successful devtools - not just as competition but as playbooks to learn from. I decided to reach out to their founders, and was lucky to find several who offered help. They shared what worked, what didn’t, and the pitfalls to avoid.

A surprising discovery while shopping for a new domain name, and a hack in the end. If you are searching for a short, one word domain name, and it is available. It might cost ~100$ to buy for one year, but renewal costs are 600$+ dollars, across all the domain registries. Seems.

Funny story.They have an app for scanning sheet music. GPT started sending them users with guitar tabs it generated, falsely claiming their product could read them. The traffic it sent pushed the team to actually build that feature.

One of the most valuable things I like to do is also the least scalable: every Rendi Basic plan subscriber gets a 30-min 1-on-1 video call with me. These chats give me raw, actionable insights straight from paying users, and they’ve shaped our roadmap more than any analytics.

About the importance of constantly talking to customers:. Yesterday I helped a paying customer with his integration, and generally with building his media app. Among the different important pointers I got - potential required features and bottlenecks in my product, he informed.

Good question about how OpenAI built a server farm for inference that doesn't burn their bank account. From the answers I've read - there's no definitive answer. KV caching is mentioned often, but I don't think it is enough to account for the scale.

A cheat sheet with everything I’ve learned about @FFmpeg.

RT @ediggs: 5 things I learned from a month of posting short form video content on @YouTube and @tiktok_us. 1) It's easy to grow the number….

We at can definitely relate. Had a tough period, but MRR is growing nicely. Might even reach break even real soon 💪.

RT @ediggs: I’m biased, but have been heavily using @get_munch over the past two weeks to created automated video clips for social media an….

This is especially seen in content creation and repurposing where the human touch should be on the storytelling and original ideas, while automation used to create the final piece of this thought process.

For sure, technology has reached a point where we can automate more sophisticated processes, which used to require manual labor and time especially cognitive thinking time and labor which is outside of scope of our main business/personal goals.

@athenascope if there are any employees looking for a job, my creator economy startup is looking for bright people from the industry - with a possibility to work remote.

RT @get_munch: We're on BetaList!.

We are at SXSW Austin, for those who are around - let's grab a coffee☕️ @get_munch .#creatoreconomy #content.

Zoom lecture (April 4) - Skilful precipitation nowcasting using deep generative models of radar with @mirowskipiotr at DeepMind, published in Nature. #ml #dl #lecture #weatherforecast #WeatherPredictions #nowcasting #gans #nature #deepmind.

RT @get_munch: We make it easy for creators to turn their content into blog posts and Twitter threads using AI.

Machine Moshe 💪.