Teiki Gen I Ep 4: Artificial Intelligence

3 min readFeb 4, 2023

Three weeks of testnet!

As we close our first three weeks of Teiki’s testnet, we are getting closer and closer to being feature-complete. The whole team is working in high spirits to hit mainnet beta by the end of the month!

Our tester community’s been scrambling, too. At the time of this writing, 80 bug bounty reports have been submitted by the community for this week alone, and we’ve already validated half of them. Let’s keep going!

(If you’re interested in joining the bug bounty campaign, hop into our Discord server! We’ve set up a pool of 20,000 ADA for bug reports, and we want to make Teiki the best playground for Creators and their communities.)

Episode 4: Artificial Intelligence

The advancement of Artificial Intelligence has been salient in the advancement of modern technology, as evidenced by creations like Midjourney and ChatGPT. Teiki integrates AI to make life easier for both Creators and Backers. Our fourth week of testnet introduces the remaining pieces to complete our first AI puzzle. Please see this earlier article for a complete overview of our integration.

Content Moderation

As a decentralised protocol, Teiki avoids censorship as much as possible. Nevertheless, unmoderated fundraising and content are potentially harmful.

We integrate moderation AI to skip manual reviews, which is a massive friction in traditional platforms. In the first generation, the content moderation AI detects inappropriate projects (for instance, raising money for political campaigns) and warns them from posting. Projects found to be violating Teiki’s guidelines will be delisted and fined their pledge.

Content not in violation but is still suggestive will appear with an advisory censor that the user may choose to display or hide with a click.

We plan to further decentralize the moderation system in the second generation by making more decisions via on-chain governance.

Text Recognition

Many projects upload information as decorated images for visual branding. Optical character recognition (OCR) identifies text from these images for analysis and search. It can extract keywords from the images to suggest additional content like taglines and a summary. Scanned text is run through moderation to ensure offensive content is not embedded.

Thanks to our OCR, the textual content of these images also helps projects be discoverable in search results. The key is to enable visually engaging content without sacrificing the power of compelling copy in graphic design.

Recommendation System

Teiki recommends projects a user may like based on their behaviours. The protocol indexer tracks these data for the recommendation system to utilize.

We currently recommend based on which project tags a user is backing. In the future, we will integrate Discord login for users to follow projects manually as another factor. We will add past interactions to the formula as well.


Niko is our AI-powered parrot mascot who guides users through several processes. For instance, Niko can detect potentially offensive content entered by the user or alert Creators to potential collaborators and competitors.

We plan to deepen these AI assistants with time, especially for creators to build and manage their communities. ChatGPT has helped create a solid trend in users interacting with intelligent assistants.

Let’s use AI to help crowdfund Creators! We’re building that future here, on our testnet!

Join Us

Website | Discord | Twitter | Telegram | GitHub