Business How generative AI can help your company build better...

How generative AI can help your company build better software


Opinions of contributing entrepreneurs are their own.

One of the challenges of building software systems and algorithms is that you often don’t have the real-world data you need to actually test before going into production or before customers start using it. It’s all too common to design a product interface or algorithm on paper, only to find that the look of the output, once put into production with real data, isn’t what you expected. GPTs like OpenAI’s GPT-4 and Anthropic’s Claude can be a game changer in these cases.

We ran into this problem Nomad data while building a new product, Data Relationship Manager, similar to a CRM for data. The product helps companies keep track of their data suppliers, datasets, purchases, interactions, meetings, tests and more. After we had a working version of the application, we realized that it was a challenge to visualize what the screens would look like in real life. We had no real user data and most of the screens were blank. This was challenging from a UI validation standpoint and also made it challenging to demonstrate the product. We wondered where we could get a meaningful amount of test data when we realized generative AI was the obvious solution.

Generative AI allowed us to do something that wasn’t possible before: generate all the usage data we needed. New generative AI models do a fantastic job with text. The key is to give them the context of what you need to create.

Nomad’s product is used by a wide variety of users in business functions. They all perform specific activities. We had to generate data to simulate a large number of user types using our product to get their work done. These activities vary in time and must occur in a logical sequence. We did this in just a few steps.

Related: I got a first look at OpenAI’s GPT-4. Here’s how it will revolutionize industries worldwide – even more than ChatGPT.

Step 1: We had to give the GPT models a general introduction to what we were trying to achieve

You are a system designed to generate actionable test data for a customer relationship management (CRM) product. Here are the steps:

First, you form a fictional management consultancy with a need for data that can be used in client projects ranging from market sizing to competitive analysis to pricing research. Come up with a very specific storyline of what specific data they are looking for and why for a number of projects.

Second, make up 10 users who work in this company. Assign any functions and titles based on the definitions below.

Step 2: We had to explain to GPT what the different users spend their time on so that it could build a realistic sequence of events

Here’s an example of one such user type that we’ll learn about in the prompt:

Data source: The employee who goes looking for data after a request from a consultant.

Role: A data sourcer specializes in finding and collecting relevant data based on what consultants ask them in response to a consulting project. They source data suppliers, initiate communication with them, ensure data quality and accuracy meet project requirements, coordinate with the consultant, and finally pass the supplier on to procurement if the consultant agrees to purchase. They log all early engagements with a data provider, such as filling out a contact form, exchanging an email, having a meeting, receiving test data, running a data test, or starting a purchasing conversation with their internal purchasing people .

Job titles: Data Sourcer, Data Researcher, Data Acquisition Specialist

We ended up learning about five different roles, but we might as well have done this for dozens.

Related: Why Entrepreneurs Should Embrace Generative AI

Step 3: We need to explain what the model should do with this information

This company records their activities around data vendors they work with and evaluates them in our CRM to keep track of everything that happened. All the work they do with the data or data supplier is recorded so that their colleagues are aware of what is happening around a data supplier and its products.

Create a series of activities for each between two years ago and today, to tell a story/dialogue about how these users interact and work with specific vendors’ data. Create activities for five to ten people for each data provider. Each user should create three to five activities for each data provider they work with.

Make sure there are activities that list experiences that actually use the data. How well did it work? Were there missing data? Was it a problem?

The output must be in a CSV format. Each row must have the format:

Date (mm/dd/YYYY), User Full Name, Data Provider Name, Data Provider ID, Activity Text


9/10/2021, Sarah Chang, AI Global Insights, sent an introductory email to AI Global Insights expressing the need for AI market data.

9/15/2021, Lisa Martin, SSC, discussed the requirements of SSC with Sarah Chang and shared a high-level overview of AI Global Insights’ data capabilities.

9/16/2021, Michael Johnson, TechIntel, has requested a subset of AI industry data from TechIntel for preliminary analysis.

Step 4: Test, tweak and test more

After running this, we noticed areas where we needed to be more specific. Within an hour, GPT-4 produced highly realistic test data:

‘2021/06/24’, ‘Emma Smith’, ‘AgriDataCorp’, ‘Contacted AgriDataCorp for an initial discussion on South American organic farming data needs.’

‘2021-06-28’, ‘John Davis’, ‘AgriDataCorp’, ‘Received AgriDataCorp data product catalog. Discussions started about costs and license agreement.’

“2021-06-30,” “Alice Williams”, “AgriDataCorp”, “First data sample received from AgriDataCorp. Started cleaning and integrating with our system.”

We were quickly able to generate an endless amount of test data, something that would have been incredibly expensive or time consuming just a few months ago.

Whether producing better products or algorithms, using GPT powered models to generate test and demo data is a must. You can bring an empty product demo to life in seconds. You can just as easily see how your products will look in the hands of real users and companies.

Related: How AI Will Transform Software Development

Shreya Christina
Shreya has been with for 3 years, writing copy for client websites, blog posts, EDMs and other mediums to engage readers and encourage action. By collaborating with clients, our SEO manager and the wider team, Shreya seeks to understand an audience before creating memorable, persuasive copy.

Latest news

Guide on Hand Painted Journal Diary Crafting and Business Insights for Artisans

In a world where digital technology often overshadows the beauty of traditional craftsmanship, the creation of handpainted journal diaries...

Essen Ladies Exotische Verführerinnen

Entdecke|Erkunde|Tauche ein in} Deutschlands top Escort-Community: Ob du nach einem stilvollen Date suchst oder einfach nur ein schnelles...

Nutte Reutlingen – Erlebe Abenteuer

Entdecke|Erkunde|Tauche ein in} Deutschlands größte Escort-Community: Ob du nach einem stilvollen Date suchst oder nur ein schnelles erotisches...

Modelle Huren Hamburg Staat – Sinnliche Verführung

Entdecke|Erkunde|Tauche ein in} Deutschlands top Escort-Community: Egal ob du nach einem stilvollen Date suchst oder nur ein schnelles...

Modelle Huren Freiburg Anziehende Damen

Entdecke|Erkunde|Tauche ein in} dem deutschen größte Escort-Community: Ob du nach einem stilvollen Date suchst oder einfach nur ein...

Nutte Reutlingen – Entdecke Verheißung

Entdecke|Erkunde|Tauche ein in} dem deutschen führende Escort-Community: Ob du nach einem stilvollen Date suchst oder einfach nur ein...

Must read

You might also likeRELATED
Recommended to you