Valeriia Kuna

This philosophy applies to every category, not just finance I think. I recently had a terrible experience with an app where I spent time answering ~30 onboarding questions. I thought I was customizing the experience, but at the very end, I was hit with a hard paywall just to enter the dashboard. It felt like they were holding my time hostage. Zero value provided, just work followed by a payment...

Definitely agree on personal finances and biometric data. But I also draw a hard line at social media autonomy. I would never give an agent write-access to my LinkedIn or X accounts to post or reply automatically. My online presence is my reputation.

I feel this! Sometimes AI efficiency feels like an illusion because the baseline for productivity has simply shifted up for everyone. Plus, the mental load of learning new tools and updates daily is a job in itself! It reminds me of the Red Queen's race in Through the Looking-Glass: It takes all the running you can do, to keep in the same place. An that's exactly how I feel every thay with the...

This is such a refreshing take in the era of automate everything! I recently watched a YC video where they explicitly advised founders to do things that don't scale in the beginning. I feel like users have developed a sixth sense for automation, they know when a founder actually spent time on them vs when it's just a clever workflow. Do you think it's even possible to build true superfans with...

Asking users to switch tabs to fill out a form usually results in a significant drop-off, so keeping everything in one player makes sense for onboarding flows! Regarding the analytics, do you provide a heatmap or drop-off analysis to see exactly which question causes viewers to stop watching, or is it just standard form completion data?

Merging the build and launch phases handles the distribution problem right from day one, which is usually the hardest part. The code export feature is a major plus for anyone worried about lock-in. When exporting the project, does the AI produce a standard stack like Next.js that is easy for a human developer to take over and scale manually?

When the system detects conflicting data points across different file types, does it flag the discrepancy for manual review or does it prioritize one source over the other?

How does the AI differentiate between a distraction and a relevant research tab when I am in focus mode? I often need to browse to work, so I am curious how strict the blocking logic is.

Congrats! How does the agent expose the logic or confidence level behind its conclusions?

I’m putting all my bets on AI for Government! The dream of AI finally killing off soul-crushing bureaucracy and processes stuck in the 90s is just too good to ignore🥹 My only worry is that most governments are still have what to do with basic digitalization, let alone AI :D But yes, it’s definitely the biggest area for a massive upgrade!

Truly love this concept, but do you think politicians are actually ready for a real-time performance review or should we expect some angry calls when their stock starts crashing? :D

@nyrelle_bade Congrats on the launch! I'm curious about the case note assistant mentioned in the comments: does it help structure the notes based on specific therapeutic frameworks, or is it more of a general summary?

Congrats on the launch! Question for a PM: once the agent does its thing, how much manual polishing of the UI or code is typically needed, or is it truly a 100% hand-off?

This is very helpful post-launch breakdown, Nika, thanks! I’m currently in the pre-launch phase and your point about Reddit really resonated with me — I’ve been feeling quite demoted lately after having a few relevant posts removed by strict moderatos. Seeing that it wasn't a major needle-mover for you makes me think about shifting my energy elsewhere. I’m curious about the strategy side — why...

The idea of having an AI team that actually focuses on the P&L and making a business profitable is a fresh take. How do you handle the trade-offs when the AI team suggests something that might be faster to build but less ideal for long-term GTM strategy?

Reteaching different AI models who I am and what I am working on is one of my biggest daily frustrations! Having one persistent brain across Claude, Gemini and ChatGPT would be a lifesaver for my workflow. Does the plugin also allow for different memory profiles if I am switching between two very different projects?

Cursor for email analogy is perfect! I love the focus on personalisation, we are actually building a personalised product ourselves, so I'm a believer that personalisation is king in this space. How does Lightfern handle the balance between using past thread context and the user’s global writing style? Does it prioritize recent interactions, or does it build a long-term tone of voice profile?

This hits close to home! We are building a personalized product in the education space (Nuomy), and we constantly see how isolated AI chats kill the learning momentum. The way you curate a feed based on past context is very inspiring! It’s exactly the kind of continuity needed to turn a simple chatbot into a real learning partner. How do you decide which topics from the past are still relevant...

Congrats on the 2nd launch! I’ve been looking at Screenpipe through the lens of a Product Manager—specifically for capturing those raw, unpolished moments during user research sessions. Having a locally searchable record of everything could be a game-changer for synthesis. Since this version focuses more on AI agent integration, how do you see it helping non-technical users (like PMs) to query...

The real shift isn't just that building software is cheaper or faster, but that the value is moving from the code itself to how well the AI understands the individual user. We are building a personalized product right now and we see that the real moat isn't the feature set, but the context and memory the AI maintains. The future of SaaS feels much more like a partnership than just a tool. How...