❜mbd

Founders

people: *create billions of onchain assets*

us: *build AI to surface the most meaningful assets to the right users*

Make web3 more personal

Not everyone is going to swift through thousands of tweets and tokens

Personalized recommendations are here for that

gm, to another day building the backbone for personalization in web3 apps, starting with Farcaster mini apps

4 steps to create your custom @farcaster social feed with @mbd:

1. Fetch candidate posts
2. Select a ranking algo 
3. Filter out the noise
4. Advanced: mix it with existing feeds for even richer content 


The four steps of our no-code feed builder:

1. Generate candidates
2. Choose ranking algo
3. Apply last filters
4. Compose with existing feeds

Guides coming soon!

Breaking: some dangerous members of the Scoppiato Faces mafia were spotted at a web3 social event yesterday. Keep your mini apps safe!
@0xcaso @bianc8 @frankk 

note: it seems like @macedo is not part of their gang (yet)

chief shitposting officer @builders-garden 🌳 • building /farville, /acidtest ⚒️

building @talent to help builders stand out 
| caretaker at /build

FarCon NYC

Been working on @mbd for 2+ years

It's really happening now. What a ride 🥲
https://x.com/YassineLanda/status/1918391383963865507

Farcaster

Proud to announce @mbd's personalization engine is powering the Coinbase Wallet feed!!

We've been working hard with the Base team these past months and couldn't have dreamed of better partners to democratize feed technologies. Still day one.

❜embed algos coming soon to all mini apps near you!

Thanks a lot to the @base for their trust! 🙏

@neynar team monitoring metrics as everyone downloads the new coinbase wallet

@emps nailed this animation for our the new @mbd website. 

Building great feeds is like making a great dish:

-> You go shopping for great ingredients: this is the candidates' generation step. Here we fetch over 50,000 data points from the last relevant posts for each user.

-> Combining many recipes for inspiration: This is the scoring step where we use an ensemble of ML models and rule-based models to score each post for the user and rerank them so that the best flavours surface to the top.

-> And finally, great presentation: This is the Visibility filtering to make sure the feed is diverse, enriched with social proof, not showing posts you already interacted with, etc.

Start cooking now by heading here: getembed.ai 🧑‍🍳

Follow to get the best of Farcaster using @mbd - AI Recommendation Engine for Web3. @a16zcrypto css23. Ex-ML Lead @Chainlink. Ex-Sr Data Scientist @AWS.

engineer /ponder, building /dash, obsessed with user experience. benbassler.com

excited to see @benbassler.eth launch this!

almost 1 year ago, during the /nounsfarcaster prop my team prototyped a similar concept using @mbd to power the feed.

unfortunately we didn’t win the grant, but seeing the idea seed grow roots by itself under careful tutelage is comforting.

I’m not completely mad.

Design

It has been and will continue to be a wild ride! I'm happy to be part of @mbd's journey as a fractional designer.

Best fractional designer on Farcaster if you ask us. Its been great working with you @emps. Looking forward to doing much more!

⚡ Introducing Dash, a Farcaster client as a mini app tailored for short-form video content.  

No onboarding needed - simply open the mini app, wait for your personalized feed to be generated and start scrolling.  

This is an MVP, so some things might not be optimized or feel perfectly refined yet. Please feel free to DC me with any feedback or bug reports!

https://dash.bnbs.dev

Check out this video mini app by @benbassler.eth

you can also use @mbd to build your custom Farcaster feed

here using a for-you feed template:

Make your app, feed or mini app feel more personal with AI powered recommendations. Incubated @a16zcrypto CSX'23 -  getembed.ai