[[Aditya Joshi]] [[Lili Feyerabend]][[Miti.ai]]

**Team**
- met at ondeck accelerator
- met to EthPrague hackathon

**Product**
- using OpenAI APIs
- there are a few foundation AI models but OpenAI is the best currently
- Salesforce has model called gen
![[Pasted image 20230104132703.png]]

**Go2market**
- protocol by protocol - they can access open data-sets before signing contracts and showcase (permissionless)
- talked to a few companies incl. Oasis
- want to start using it as an internal tool
- They are using the BigQuery Ethereum data set
- They are building the data pipeline from scratch including the parser etc.
- They have decided to build it themselves because.. they want to reliability
- They want to use OpenAI to parser writer
The traction:
- They have engaged several companies
- Right now they think the best way is to go protocol by protocol
- They focus on public/private data sets
- For companies they could automate the data scientist
- They are thinking about a public facing product (scoped version to growth hack

**Competition**
- OpenAI basically has a monopoly
- companies will try to move to open-source alternatives with own NLP models
- hard to replicate

web2 comparables
- [https://thoughtbot.com/](https://thoughtbot.com/)

**Round**
- raising between 1.5-2M