Gartner recently came out with a research paper on AI applications in organizations and they talk about Game Changing AI and Everyday AI. While we have all heard examples of Game Changing AI (Vodafone, Klarna, Gong) it requires a lot of technical expertise and is quite expensive. It is game changing and well worth the investment.
That being said, what are the options for folks that want Everyday AI - It is to use ChatGPT or another LLM? There are quite a few drawbacks with this approach -
It is not secure, so you don't want to upload your data.
The interface is a chatbot and requires you to be asking questions and responding real time.
You cannot combine a series of prompts into a workflow
You cannot repeat the same workflow multiple times.
You cannot pick the LLM that is best for the right task.
I have started working with Dmitri Tcherevik and Helen Fu Thomas at AnyQuest and playing around with the AnyQuest Builder and have been having a lot of fun. Reminds me of the fun times I had building dashboards with Tableau (in the early days) when you would lose sense of time as you were so in the flow of the analysis.
AnyQuest basically lets you string together a series of prompts in a workflow (you can execute tasks in parallel or serial) that you can run on a set of data. The example, I was playing around with was to do research on some private companies. I had 3 companies and I wanted to find out stuff that I would normally have Googled - what does the company do, where are they located, how many employees do they have, who are their execs, have the execs spoken on any podcasts, who are their competitors, what are the priorities of their nearest public competitor etc. Again the questions themselves don't matter, the point is that I can spend hours doing this and I wanted to see if I could automate this with GenAI and AnyQuest did that for me.
Here is the link to the demo - https://youtu.be/B1hIbYUN-eM
I gave it 3 company names -
Company Name : Anomalo, MongoDB, Retention.com and asked it to generate all this information.
Here is a link to the output - https://drive.google.com/file/d/1GgPxIVKLYyKtODk5vb_EeWDtJywnACO7/view?usp=sharing
Some points to make -
The specific use case does not matter. If you are doing any kind of Google search on any topic and you do a series of searches and you repeat them every time, it can be automated.
There are some really good point solutions for account research (Salesmotion.io) and that will meet the need of a lot of folks. This meets a different need.
AnyQuest is extensible and can call APIs to pull information from systems. So it is not just about Google Searches
It is secure and can run within your environment or in a secure way.
You can also upload use cases / documents etc. to target the information you want back.
I can make the output much better with iteration (I am sharing the raw V1 output). It is a lot like building a Tableau dashboard.
Curious if people have similar use cases? What do you think about the need for a Tableau for GenAI?
Jeremy Blaney Mike Madigan Dmitri Tcherevik Helen Fu Thomas Anand Ekambaram Eepsit Tiwari Anoop Shah Abhijeet Khadilkar
