Since ChatGpt landed in November 2022 the world seems to have gone Generative AI crazy. It’s pretty cool how viral tech like this rapidly becomes the number one thing to talk about. The question we get asked most is “how can I actually use GenAI in my business?”
A lot of people see the potential but converting this to business application is where it becomes more difficult. So through this post we’ll share some ideas of how GenAI can be effectively put to use.
Wow, this is amazing
Like a lot of tech it’s easy to get excited and expect too much too soon. This is what’s happening with GenAI, resulting in failed projects and a dip in business trust. For a lot of business users their experience has been underwhelming.
We think it’s great so stick with us and we will explain why.
If we look at Chat GPT in our personal lives it really can be transformative. These example use cases are where it largely outperforms the old way of using a search engine.
Create me a great advert to sell a chesterfield sofa in under 50 words Write a party invitation for my sons 8th birthday party How do you explain gravity to a 7 year old Which football manager has the most wins and losses in the Premier league between 2015 - 2020
The answers you get from the above questions are way better than if you try the same via Google. The great thing is you can also keep asking Chat GPT for more clarity, subtle changes and it will iterate the answers until you have what you want. All sounds great, right?
Over delivering at home but underwhelming at work
However, once you start to deploy a tool like this at work, a few stages of disillusionment appear.
How do we train a model that knows nothing about my business? Feeding it our data for training sounds like a security risk? Now that it has some of our data, it still struggles to make sense of it. It’s missing the logic required.
But all is not lost, GenAI really will be transformational to your business
Having spoken to a lot of business leaders trying to implement GenAI, there are 3 key areas the tech tends to fall down:
- Security
- Business logic
- A predictive capability
That’s not to say we give up and move on. The potential of the tech is too great for that. We just need to find a way to address these concerns. Imagine a world where you can securely talk to your data and it understands your business logic and processes, whilst giving you answers about the future. That sounds like a recipe for success!
You could use this to get answers to real world business questions, such as; which of my customers will churn next month, which of my leads are most likely to convert, how much stock do I need to hold to reach demand, and how many drivers do I need next month?
The good news is, all of this is currently possible.
Option 1
You could build the logic, security and predictive capabilities yourself. This approach has been taken by a number of organisations but absolutely requires in-house data science and engineering teams with free cycles to build the bespoke tech. Thinking long term, these same teams then need to run, maintain and continuously optimise the models. So best hope that engineer who knows how it all works doesn’t leave!
Option 2
Alternatively, you can licence SaaS models and get someone else to do all this hard work for you. Allowing you to benefit from the vendor's years of skills and investment and their ongoing commitment to making their product better. The best vendors offer out the box solutions that will build custom models for you on the fly. So you get the best of both worlds, low investment (time and money), and custom models with built in logic and security.
Shameless plug
At Infer we’ve built a platform that provides the SaaS capability mentioned above. And if you were wondering what it does…
Our GenAI securely creates bespoke AI models on the fly which surface predictive patterns and insights whilst also creating suggestions on how to optimise your particular use case or KPI.
The GenAI coworker will take 30-60 seconds to scan your data, provide suggestions on what it can do with the data, build a custom AI model and then tell you all underlying factors, patterns and create predictive models.
For example, if you work in logistics, in 60 seconds you go from having a huge, messy data set to having predictive models on optimising on time delivery and knowing which factors are most impacting your delivery time.
All this information is surfaced through graphs and explanatory text so you can share it with the wider business.
In Summary
GenAI will be transformational to business and is already having a positive impact on growth and profitability. Out of the box it’s almost useless as you need to build applications and security layers to give it relevance to your business. You can either build these applications yourself or buy them pre-built from companies like ours and hundreds of others. But if you like the sound of what we are doing, here’s a 60 Second Infer Walkthrough
Good luck with your GenAI plans.
