Hey all. I'm really interested in how you evaluate your agents in production.

The standard approach is sample, LLM as judge and possibly a golden dataset. That seems to be the standard setup.

Last month, we finally got a demo with a massive prospect a 500-person company that was exactly our target audience. The demo went perfectly. The VP of Product loved our tool, and the team was ready to sign a $1,500/month contract.

Then, the deal moved to their IT and Security department.

Remember when products designed for consumers unexpectedly revolutionized how businesses operate? For example, Discord started as a chat app for gamers but is now essential for remote teams.

Its journey from gaming-focused platform to business tool is fascinating:

• Originally built for gamers needing low-latency communication

• Features like server organization, voice channels, and role permissions were perfect for gaming

• Discord thrived during the pandemic as businesses discovered its community-building strengths worked brilliantly for remote collaboration

• Now powers everything from customer communities to virtual events

I think by 2030, the real question won't be "Do you use AI?" but "Can you work without it?"

I noticed this in my own life recently. 

I always said "I'm terrible at design" whenever I needed to make something look nice. 

Finding interested people is one thing. Finding users who have the exact problem, will try an early product, and give useful feedback is much harder.

Where should a developer actually look for those first 100 real users via product hunt?

I bought a pair of wireless earbuds a while back with 4.5 stars and thousands of reviews. Sound was great, battery was great, everyone loved them. Around month 8 the left bud started cutting out, and by month 10 it was dead. Warranty had ended two months earlier, of course...

When I went back to the reviews, I realized almost all of them were written within the first few days. Amazing sound! Love the case! Nobody comes back after 10 months to update their review, so the rating mostly tells you how the product feels out of the box, not whether it survives...

Now I scroll straight to the 1-star and 2-star reviews and sort by oldest or search for words like months later, stopped working, and after a year. That helps a bit, but it s a lot of effort for something I m just trying to buy.

It also feels like this applies to everything: backpacks where the zipper gives up, water bottles that start smelling weird, standing desks that begin wobbling. The stuff that matters most is exactly what a launch-week review can t tell you...

most audit trails answer 'did this action happen.' the ones that hold up under legal scrutiny answer something harder: 'would the reviewer who signed it still say yes if you asked them today.'

ive been calling this consent-of-record. not just reviewer id, but reviewer knew X at time of sign-off and stands behind it now. it sits above the usual forensic level (timestamp, user, action, immutable hash chain). that layer proves it happened. it doesnt prove the human is still endorsing it.

There are plenty of tools that seem useful on paper, yet something makes us close the tab instantly.

Sometimes it s pricing, sometimes messaging, sometimes the onboarding feels unclear.

Often it s a gut feeling that s hard to explain.

When I launched my first paid tool, I set the price at $5 a month because anything higher felt like I d scare people away. It took me almost a year to realize the people paying $5 were the ones emailing support the most, and the few who would have happily paid $30 never thought it was a serious product.

I raised prices earlier this year and lost a couple of users, but revenue went up and, oddly, the complaints went down. I still feel a little guilty every time I look at the pricing page, though.

I ve noticed two patterns with makers I talk to:

1. People who start cheap to get traction and then struggle for months to raise prices without upsetting early users.

2. People who charge more from day one, get fewer signups, but end up with customers who actually stick around.

Hey PH community! In two weeks, I m launching my app, SUN, here on Product Hunt. It s been a journey, and I m now focused on how to get meaningful visibility and upvotes. I ve been digging into what others have done and came across these 5 common strategies:

1. DM outreach on LinkedIn

2. Joining Product Hunt-focused WhatsApp groups

3. Engaging in Facebook groups for PH launches

4. Running social media ads that link directly to the PH page

5. Buying Product Hunt ads (starting at $5,000 )

Has anyone here tried any of these?

What worked? What felt like a waste of time or money?