GenAI for building a case study (synthetic data)

People keep asking for practical examples of GenAI and my approach has always been about applying technology to solve real business needs. One of the projects I am working on is to build a GenAI for Sales program for upGrad . upGrad was formed to impact the lives of working professionals by helping them upskill while they work and as we all know GenAI is going to quite disruptive and we need a way to upskill people while they work.

It's been a fun 2 months as I've gone about distilling my experience in how Sales leaders can apply GenAI in practice and building a curriculum that goes past the theory, goes past prompt engineering and gets people hands on with the various options available to them. We look at point solutions, we look at LLM based options, we look at features within existing products and we also get hands on building a few agents. You can hear about GenAI for Sales on my podcast with Srikanth Iyengar here - GenAIrous. .

To get practical and hands on with workflows that sales can relate to I had to come up with a fictional company they worked for, a fictional company they could sell to, fictional leader profiles they could sell to, fictional company information etc. Here is a list of the things I needed to come up with -

1. Name of company they worked at.
2. Company overview, background and product offerings
3. Case studies and testimonials
4. Name of company they wanted to sell to
5. Overview, background and product offerings of the buying company
6. Org structure of the buying company
7. LinkedIn profiles for 2 executives
8. LinkedIn posts for 2 executives
9. 2 blog posts for each executive
10. 2 news articles about the executive
11. A podcast / youtube video for each executive
12. A call transcript of a sales conversation

Overall about 50 pages of content and this would have taken me 100+ hours. Initially I was planning to use real profiles and companies but that was not a viable option. So I started experimenting with using GenAI to generate all this content and I was able to create everything except for the youtube video.

To keep expectations in check, it is not magic. I had to come up with the right prompts, check the outputs and make sure that the story flowed, amde sure there was variety etc. It probably still took me 20-30 hours to get a finished product but it was so much quicker that me having to write this all from scratch. Also, I am not the best writer and I have a specific writing style that an old boss called 'bullet formy'. GenAI helped me go from being a writer to an editor and this I am good at, as I know what I want in the content.

Here are some examples -

LinkedIn Profile - Charlotte Green .

LinkedIn Posts - Charlotte Green .

Blog post - Power of empathy .

News Article - CS Leader of the year

I still see so many articles about GenAI being hype and the problem is that people are treating it like it is a mature technology. In all technology shifts in the past, it used to be early adopters who helped the technology mature (think about the first Tesla, the first iPhone etc.). With GenAI enterprises have gone all in due to FOMO and it does not align with the enterprise mindset.

That being said, they have all gone all in and if they want to make the most of the investment, they are going to need to invest in Education and upskilling and programs like what upGrad is building will be key to getting the most value from GenAI.

upGrad Srikanth Iyengar Shreyasi Singh Himanshu Gupta