Welcome to Yapmandu !!

No-noise catchup on your Twitter Tribe

What did I miss?

Influencers at your fingertips

Your engaged community speaks

Last 48 hours, ranked

Why This?

First of all, I am not affiliated with the organizations I choose to include here. This is just a personal side project.

I love Twitter, but it can be noisy and you don't automatically see everything of interest. You could use Tweetdeck or Lists, but frankly I'm not a fan.

So, for a super-fast & focused catchup, this is for you 👊

I know ... yapmandu is an odd name - what does it mean?

This started as just playing with the new V2 Twitter API, and ended up going down this rabbit hole. Fun project & it may become a training class on how to build a lightweight serverless webapp with Twitter API integration. TBD.

Feel free to send me your enhancement requests 💡

How it Works

I picked the following Twitter tribes to kick this off, since I have a general interest in checking in on them all. Following are the links to access:

A Tribe generally consists of (1) VIPs (2) Hashtags (3) Engaged Community - Fans!

Behind the scenes, the Twitter API is periodically accessed to pull (1) Tweets from specific users (VIPs) (2) Tweets containing the go-to Hashtags ... the look-back period is 48 hours. The tweets are ranked by engagement and presented as a max of (1) top 50 tweets per VIP and (2) top 150 tweets by the community.

Example setup for CRYPTOBEASTS:
VIPs: @jack @michael_saylor @cz_binance @RaoulGMI @LynAldenContact @nic__carter @APompliano @laurashin @gladstein @jackmallers @Melt_Dem @BitcoinMagazine @100trillionUSD @woonomic @coinbureau @lopp @PeterMcCormack @Breedlove22 @WClementeIII

Example setup for REAL VISION:
VIPs: @RaoulGMI @Remi_Tetot @AshBennington @JackFarley96 @edwardnh @JulianMI2 @RealVisionBot
HashTags: #RealVision #ExponentialAge #FestivalofLearning #RVLearning

Totally open to suggestions about who to include & relevant hashtags. The data is refreshed every few hours, ie. not real-time. Why? Because there are Twitter rate limits on how many tweets can be pulled over a 30 day period. This is built with the new TwitterV2 API - as soon as Twitter offers more rate options, these limitations can be lifted. In the meantime, enjoy the quality tweets being presented - they are ranked by engagement, so there should be some good tweets to check out.

The ranking system is based on a weighting of all 4 metrics buckets - Likes, Comments, Retweets, Comment Retweets. Least weight for Likes, most weight for comment retweets. The more buckets have values, the better the rank. For example, having 1000 retweets ONLY, may not win out out over a lower # in all 4 buckets.

Regarding TwitterV2 API Integration, the top tweets over the past 48 hours are pulled each time, along with (1) Associated media (2) Referenced tweets (3) Referenced Links etc. The idea is relevant recency in order to somewhat satisfy what did I miss?!

NOTE: Retweets from the community/fans are excluded since the Twitter API rarely produces correct metrics for retweets

Best way to get included here is to bring your own API Key (Bearer Token). With that, the rate limits are spread out ... more access for everyone.

About

I'm Paul Heath, semi-retired 2x CTO co-founder from the travel-tech world (nothing you've heard of). After 30+ years of countless projects and co-founding 2 companies, I figured that was enough brain damage for a while.

During Covid I've been working on my cooking & cocktail game, learning about crypto, and coding for fun.

If you find any bugs, I blame the Manhattans 😉

The Tech

The approach was to leverage AWS serverless infrastructure and minimize code to maintain. Following are the bits ...

Front-end

  • Plain ole HTML, CSS, Javascript
  • No Javascript Frameworks
  • Included tweets are searchable
  • Fontawesome
  • Google Font (Karla)

Middle Tier

  • AWS Lambda
  • AWS HTTP API Service
  • AWS Rekognition (Text from Images - Premium)
  • AWS EventBridge (Scheduler)
  • AWS SNS (Alerts)
  • Javascript/Node.js
  • API to Twitter V2

Database

  • AWS DynamoDB
  • Store Site Profiles
  • Compress & Cache the Ranked Tweets

Deployment

  • AWS S3
  • AWS Cloudfront
  • AWS Route53 & Certificate Manager
  • Progressive Webapp (PWA)
  • Offline Viewing