stefan_r

Wave | Applied AI Scientist | https://wave.com | Remote (US/UK/Canada/Spain) | Full Time

Wave (YC W12) is making Africa the first cashless continent by building affordable, instant financial services for millions of users. We’re the largest financial institution in Senegal and Côte d’Ivoire and are growing rapidly across 9 countries.

We are looking for an Applied AI Scientist with 5+ years of experience and deep expertise in building, evaluating, and optimizing autonomous agents. You'll work on autonomous voice and digital agents powering 10M+ interactions monthly and will own features end to end, tackling hard edge cases like low-resource languages, low literacy, and poor connectivity.

Salary up to $227,900 + equity. Benefits include 6 months parental leave, $10k charity match.

Apply here: https://www.wave.com/en/careers/job/5779974004/

Got it!

Can't find any submission with gab18 in it sorry! Need some more info like former employer, initials, something likely to be unique in your resume or submission, or exactly what you put in "how did you hear about this job"

Not quite a referral but put gab18 / hacker news on your application and feed back here once it's posted and I'll make sure it's looked at!

Wave | Applied AI Engineer | https://wave.com | Remote (US/UK/Canada/Spain) | Full Time

Wave is making Africa the first cashless continent by building affordable, instant financial services for millions of users. We’re the largest financial institution in Senegal and Côte d’Ivoire and are growing rapidly across 9 countries.

We are looking for a product-minded Applied AI Engineer with 5+ years of experience and deep expertise in building autonomous agents. Experience with Voice AI or at vertical/coding agent companies is a major bonus. You'll build autonomous voice and digital agents powering 10M+ interactions monthly and will own features end to end, tackling hard edge cases like low-resource languages, low literacy, and poor connectivity.

Salary up to $222,700 + equity. Benefits include 6 months parental leave, $10k charity match.

Apply here: https://www.wave.com/en/careers/job/5779974004/

We used ngrx/store for state management, the main redux implementation for Angular 2. Having centralized state was very powerful, helping us implement more features in less time.

In non-HTML5 browsers you need # for routing single page applications.

The 30 second tracks come from spotify. Their API doesn't provide for a way to get the full track even for authenticated users, so we use an iframe widget for that :/

Go to the detail page by clicking the tiles or track titles, and look at related artists/tracks :)

Spotify also has pretty good recommendations, but requires people to be authenticated to use their recommendations API. Our recommendations currently come from Last.fm.

Free. Auth0 is definitely very cool, makes it easy to implement Github/Facebook authentication (or any other service) as well. A bit of a pain to implement in Angular 2 right now if using hashes for routing (see https://github.com/auth0/lock/pull/790#issuecomment-27426628...), so we used popups as a temporary workaround.

We debated putting YouTube music in the search results but decided against it for now, some of the coolest features involve the built in audio player.

Thanks for the feedback! This being Angular 2 we actually used TypeScript, with some Django on the backend (with some Node.js and Ruby for generating the waveform data using ImageMagick). We also used Three.js for the 3D visualizations -- the sphere one is especially cool. Put on on some music and check it out by clicking the menu and then the circle icon on the top left.

We also have some cool 2D visualizations in the equalizer (click the arrow on the left of the player to access).

All in all this was a very fun stack to develop on, once we had all the initial pieces in place (getting ngrx/store set up mostly) we added new features very quickly. We got to where we are right now over the course of less than 3 weeks.

This is all hosted on a tiny DigitalOcean droplet and we use external APIs with daily limits, so hoping it all stays up while it's on the front page here!

We use the Last.fm API for music discovery through the track detail pages, you can get there by clicking the tiles or the track titles.

Openwhyd looks very cool!

Our top tracks come from SoundCloud and search gets tracks from the SoundCloud and Spotify APIs -- detail pages contain links to related artists and tracks.