Comments URL: https://news.ycombinator.com/item?id=42795082

Points: 1

# Comments: 0

Comments URL: https://news.ycombinator.com/item?id=42172079

Points: 2

# Comments: 0

I wanted to share what I've been building for a few months now.

Playrun https://playrun.dev

Playrun is an AI-first QA platform. As an engineering team, ensuring nothing break on every production release either requires the team to manually QA or setup CI pipelines and test automations that are painful to maintain. - I was not happy with these option.

So I built Playrun - Our approach was to build a multi-agent AI, that maps out your application flows like a user, generate test for each flow and allow you to setup a schedule to run generated tests.

I've personally been using this on a small team of 4 at my 9-5 for about 2 month now and it has significantly reduces how much regressions and bugs gets into production.

We'll love to hear your thoughts especially small teams like us who need to move fast without breaking stuff. Thank you.

Comments URL: https://news.ycombinator.com/item?id=42136157

Points: 1

# Comments: 0

Comments URL: https://news.ycombinator.com/item?id=41995425

Points: 1

# Comments: 0

Comments URL: https://news.ycombinator.com/item?id=41698505

Points: 2

# Comments: 0

I recently worked on a project rewrite with a small team on a tight deadline. We needed to keep shipping while also writing tests for the application. We believed unit tests were a waste of time and decided e2e tests were the way to go, but doing so manually was still time-consuming and made shipping slower.

Testing is one of those things we do as engineers that don’t have immediate ROI, and it was almost impossible to make the business team understand why we needed to do it and why we were spending so much time on it.

Augmenting testing with AI helps reduce the dev hours we invest in testing so we can focus on shipping, which is very important to the business and users.

Basically, playrun.dev crawls through your application and discovers possible user flows. Once a flow is found, it is mapped step by step, and each step is run and evaluated to ensure it’s correct before the next step is mapped until it can be determined that the flow has ended.

Under the hood, we use multi-agent LLM to discover flows and generate tests.

All that is needed is to provide the public URL to a testing version of your application (please don’t use production URLs, lots of dummy data will be generated). We will navigate the whole application and generate Playwright tests. We also provide a CLI tool to export generated tests directly to your Playwright project.

Although this isn’t perfect (we still had to review and manually fix a few tests), it was able to generate almost perfect tests for about 70% - 80% of flows in our applications (>90% on some runs, we hope to reproduce that consistently).

We still have some limitations and gotchas around Captchas and 3rd party authentication (Google and some other providers block testing browsers as bots and the only way we are able to get around Captchas for now is by disabling it on our testing server).

Additionally we are working on ability to generate, manage and run you e2e tests on a schedule in one place. So regressions can be detected faster on autopilot and if application flows changes, they can be easily be remapped without writing a line of code.

You can try it today at https://playrun.dev.

I’m really happy to show this to you all, thanks for reading about it, please let me know your thoughts and questions in the comment.

Comments URL: https://news.ycombinator.com/item?id=41648961

Points: 5

# Comments: 0

Comments URL: https://news.ycombinator.com/item?id=39764273

Points: 2

# Comments: 0

Comments URL: https://news.ycombinator.com/item?id=39383656

Points: 2

# Comments: 0

Comments URL: https://news.ycombinator.com/item?id=38752273

Points: 3

# Comments: 0

With Diarupt AI, you can build:

Dynamic video and audio-based AI interactions, with human-like avatars and realistic voices, that can have long and fluid conversations, exposed through easy-to-use APIs & SDKs so you can integrate in minutes.

At Diarupt, we believe user experience is an essential part of any technology, and the same is true for AI. We aim to take our everyday experience and interaction with AI to the next level by making it seamless and as natural as possible.

Diarupt is in Public Alpha, and we will be looking forward to your feedback as we build the future of AI Interactions together.

We’ll be granting access in batches over the coming weeks, so join the waitlist and watch out for your invite.

We are so excited to see what you’ll build.

Cheers

Useful links:

- Website: https://diarupt.ai

- Demo: https://youtu.be/eROf7SCdqcQ?si=B2nyqxBeTssGN-3E

- Quickstart: https://docs.diarupt.ai/quickstart

Comments URL: https://news.ycombinator.com/item?id=37687570

Points: 5

# Comments: 0

Comments URL: https://news.ycombinator.com/item?id=37673307

Points: 3

# Comments: 0

Comments URL: https://news.ycombinator.com/item?id=37658050

Points: 5

# Comments: 0

Comments URL: https://news.ycombinator.com/item?id=37642375

Points: 6

# Comments: 0

> Is it really as fluid as shown in the video, or are there pauses while the text and audio/video are generated?

Yes the video is as smooth as shown in the video, and there are no pauses while video is generating.

There is alot of moving parts to get it be that fluid but for the video generation we built a NeRF-based framework which allows for realtime synthesis.

> Do they wrap something like gpt-3.5 and HeyGen or D-ID APIs? Did they build their own system?

We didn't use HeyGen or D-ID, we tried these out initially but when we plugged in all the moving parts they were too slow to allow for realtime AI conversations, which drove us to build out our own system. On the LLM side we use GPT-3.5/4 as the brain which is currently probably the most efficient (cost/output wise) at the moment but we also have Llama 2 integration in the works.

> Is there a paper or even a GitHub repo with these models out there somewhere to build off of by any chance?

We're building this out as an API, you can integrate into your own product, to build more complex flows.

It's still in private alpha and we are still polishing the edges but you can join the waitlist (https://diarupt.ai/) to try it out once we start rolling out.