Announcing Digest

[Brian Muhia]

Hey guys,

I just wanted to do a shameless plug of my startup with Adrian. Our product Digest allows researchers to quickly answer questions from other papers, find relations between disparate sources of text and extract specific data from large sets of documents. All this by simply asking the question you want answered.

It helps lawyers do case research and quickly sift through information for legal due diligence. Want to find out how many lawsuits the company you’re researching has gone through? Just upload all their documents to Digest then ask ‘What is this company’s policy on data privacy?’

CEOs and other executives can use it to improve the speed and quality of their decisions, students can use it to speed up research and learning (including using our amplification feature to generate questions from uploaded documents to test themselves) and hopefully it will help in many other use cases that we haven't even thought about let alone seen. 

The only question here is ‘What do you want to know?’

Pricing starts at 3 USD for the basic Chanuka account, this is meant for light use as on average this might allow for ~100 questions (this varies based on file size and amount of text used by the AI model for context). This account was primarily made with students in mind, just upload any softcopy notes/texts from your courses and you have one access point for everything you need to know. 

The other individual package goes for 15 USD and simply gets you more documents and more questions. Credits that haven’t been utilized by the end of the subscription period get rolled over to the next month.

If you’re a knowledge worker overwhelmed with documents and research, you’re interested in exploring a locally made AI product and you want to check out an AI product built with AI safety in mind head on over to https://digest.fahamuai.com.

PS Even if you’re only registering to tear our system apart in a tech review, we welcome your input, and the cash it comes with.

I am working on a paper to explain this system and others like it. Here is the Google Doc if you want to have a look:

https://docs.google.com/document/d/160Yw_iuvztB6CTT9Osj5wC0sOrEKjfaGkkeeYuwQf4Y/edit?usp=sharing
Kind regards,

Brian

[Benson Muite]

> https://digest.fahamuai.com <https://digest.fahamuai.com>.

>
> PS Even if you’re only registering to tear our system apart in a tech
> review, we welcome your input, and the cash it comes with.
>
> I am working on a paper to explain this system and others like it. Here
> is the Google Doc if you want to have a look:
>

> https://docs.google.com/document/d/160Yw_iuvztB6CTT9Osj5wC0sOrEKjfaGkkeeYuwQf4Y/edit?usp=sharing <https://docs.google.com/document/d/160Yw_iuvztB6CTT9Osj5wC0sOrEKjfaGkkeeYuwQf4Y/edit?usp=sharing>
Thanks for explaining the system. Perhaps put the document in a code
repository and use Markdown/Asciidoc/Org/Jupyter notebook, see for example:
https://www.ncbi.nlm.nih.gov/pmc/articles/PMC7359270/
Maybe:
https://github.com/kedz/newsblaster
might also be of interest, though has not been updated in some time.
> Kind regards,
> Brian
>
> --

[Brian Muhia]

Thank you Benson, I put a Colab/Jupyter notebook in this GH issue, which also has a link to a 30minute talk I gave about the work: https://github.com/poppingtonic/transformer-visualization/issues/12

Brian

--
You received this message because you are subscribed to the Google Groups "Nairobi GNU/Linux User Group" group.
To unsubscribe from this group and stop receiving emails from it, send an email to nairobi-gnu...@googlegroups.com.
To view this discussion on the web visit https://groups.google.com/d/msgid/nairobi-gnu/e7773f24-619e-5981-0ed0-0bf53a17173e%40emailplus.org.