Using AI to boost efficiency and creativity in business operations

ChatGPT has been making headlines since launch given the seemingly broad applicability of its large language model (LLM). The model was trained on vast amounts of data, which inevitably included business operations context. 

‍

One of our main aims with Flawless is to empower business operations teams to increase their efficiency and effectiveness by working better with data (& thus drive more impact for the business). One of the main challenges, as any ops manager would tell you, is that said data is extremely heterogeneous, thus hard to standardize (even across companies from the same sub-vertical). 

‍

So when ChatGPT came online, we naturally started testing different ways in which it could help ops teams. There is a lot to be said about the broad applicability of Generative AI - the possible use cases are endless, especially if we factor in (the likely) exponential development in the next months / years. For now, we’ll keep it simple and explore 2 areas in this article: manipulate data (increase efficiency) and discover hidden insights (boost creativity). Both are crucial - as data volumes grow exponentially, traditional methods of “managing your business in a data driven way” are becoming increasingly inefficient (think of creating a dashboard for every business question). 

‍

How to manipulate your data 10x faster

‍

Extracting, working with and keeping data fresh is not trivial. We spent countless hours thinking how to improve ways of working between data and business teams. Usually it boiled down to optimizing back and forth between the 2 teams on creating / running SQL queries for the data the business teams needed. So, imagine business teams could do that themselves. ChatGPT enables this - with the right prompt, it can generate and explain the required SQL.

‍

Let’s explore a concrete example - let’s assume that you want to get a list of vendors that have declining performance and above average volumes. 

‍

Notice that ChatGPT assumed table names and columns that it would use, which likely means the query won’t execute for your environment. So let's give it the context to enable that. 

‍

Notice it now uses the correct table and column names, which should make this query usable. Needless to say, we’ve assumed little to no SQL knowledge for the above. However, data teams can also use this, while likely cutting 50%+ of the workload required for SQL data manipulation tasks. 

‍

At Flawless, we’ve taken this approach a step further. Given our clients have their data connected, we are able to offer quite accurate Text to SQL capabilities on top of the required data. This has already saved hours (from both data and business teams) and enabled new use cases within the platform (see it below in action on the aforementioned use case). 

‍

How to discover hidden insights 

‍

The second area we wanted to explore is how ChatGPT could help business users brainstorm for their specific sub-vertical / problems. Again, let’s assume a scenario where you manage a real-estate classifieds platform and are struggling with 2 specific issues. 

‍

For someone working in classifieds, this answer is likely stating the obvious. However, let’s improve the prompt and ask for specific deep-dives:

‍

This is starting to look much more actionable - while you might already do a few things from the above, this list will likely spark interesting conversations within your management team on where things could be improved regarding ads moderation.  

‍

To conclude - experiment, experiment, experiment

‍

ChatGPT is an invaluable tool for business operations teams looking to boost their creativity and efficiency. With its advanced understanding of SQL and vast knowledge base it can act as an effective brainstorming partner for business teams. The key is to start using it and experiment with different inputs, in order to understand how to make the most out of it. 

‍

If you’re interested in trying out Flawless, book a call with us here.