This is a much better way to evaluate these platforms than just comparing feature lists. I’d rather see actual resolution rates, escalation rates, and maintenance effort from a team that has been running one for six months.

Chatbase and Voiceflow stand out to me for different reasons. Voiceflow seems more focused on designing the conversation experience, while Chatbase seems to be moving more toward actually running customer support workflows.

One thing I’d add is that agents can make processes more visible too. When you have clear instructions, actions, and logs, it becomes easier to see where something went wrong instead of just finding out later that a task was never completed. That’s a pretty underrated benefit.

‎I’ve started looking at SaaS differently because of this. Features still matter, but I’m increasingly asking whether the product reduces the amount of work I have to do. If an agent can genuinely take something off my plate without creating more work to supervise it, that’s where it gets really valuable.

Honestly, I’d look at customer support. There’s still a lot of repetitive work around tickets, refunds, order updates, and follow ups that AI could handle while still handing complicated cases to a human.

I like the concept, especially for teams that are constantly tweaking their agents. Once an AI agent is live, there’s always something to review or adjust. Having one place to do most of that could make the whole process less tedious.

For me, one agent feels better from the user’s perspective, while multiple agents might be better behind the scenes. The customer shouldn’t necessarily need to know how many agents are working underneath.

I think the biggest mistake is trying too many channels at once. Getting the first 100 users is probably more about finding where your actual users already hang out and talking to them consistently. For some products that might be Reddit, for others it could just be direct outreach and referrals.

I’d probably lean toward AI-native, but I wouldn’t switch just because the AI sounds better on paper. What matters to me is whether it actually reduces the amount of work the support team has to do. If humans still have to clean up every AI response, the “AI-first” part doesn’t mean much.

I’ve started thinking about this the same way. Customization is great when it solves a real problem, but it can quietly turn into technical debt. I’d rather keep the core AI setup simple and only add custom logic where it actually improves the customer experience.

Honestly, I think launching early makes more sense for most SaaS products. A waitlist can build hype, but real users will show you things you probably wouldn’t catch beforehand. I’d rather have 20 people actually using the product and giving feedback than 500 people waiting for it.

I like the idea, especially because managing different bots for every channel can get messy pretty quickly. But I think the shared context is the part that really matters. If the agent can remember what happened across channels and still hand things over to a human properly, that’s where I’d see the biggest value.

For me, it’s the constant AI suggestions that get annoying. Sometimes I’m just trying to finish a simple task and the product keeps trying to help when there’s really nothing to help with. I’d rather have AI quietly handle the boring parts in the background than have it interrupt me every few seconds. The best AI features should feel…

Read more

I’ve used PocketBase for smaller projects and honestly, the simplicity is what makes it so appealing. But I’d also be a little cautious about making it the foundation of a SaaS that I expect to grow a lot. The MVP stage can hide a lot of backend problems that only show up once users, data, and edge cases start piling up. Would love to hear… Read more

I’ve been wondering about this too. The live Shopify data is probably the part that makes the biggest difference for me. There’s a big gap between an AI saying “check your email” and actually being able to tell you where your order is. I’d be curious to see how reliable it is with real customer orders though, especially when things get a bit… Read more

Honestly, I think that’s where things are heading. I used to look at AI support mostly as a faster way to answer customers, but now I’m more interested in whether it can actually solve the problem without someone having to step in. I still wouldn’t want to get rid of human support completely though. Sometimes you just want to talk to a real… Read more

I think the biggest difference is that AI agents can actually move the conversation forward instead of just waiting for someone to ask a question. I’ve seen them work well for things like product recommendations, pricing questions, and lead qualification. But I wouldn’t completely remove live chat yet. For anything sensitive or high-value,… Read more

I’ve noticed the same thing. A lot of the SaaS products I find most useful aren’t the ones constantly being talked about on LinkedIn or Product Hunt. They’re usually smaller tools that solve one specific problem without making the workflow more complicated. I’d actually love to see recommendations from people who’ve used these products… Read more

This is honestly the kind of SaaS list I enjoy more than the usual big-name recommendations. Some of the smaller tools end up being way more useful because they solve one annoying problem really well instead of trying to do everything. I’d definitely be interested in seeing what lesser-known tools people have discovered through actual day-to-day use.