Transforming business with Generative AI and AWS – a case study
The public release of OpenAI’s ChatGPT has opened the floodgates for Generative AI around the world with many other companies following suit. Often referred to as a ‘game-changer’, many expert analysts believe that AI will usher in the fourth industrial revolution, touching every part of our personal and professional lives completely.
‘A.I. is not all hype. It’s the “fourth industrial revolution playing out”’.Dan IvesWedbush Analyst
Introduction
Generative AI and LLMs (large language models) have taken the world by storm in recent times.
Generative AI is a highly innovative field that enables machine learning algorithms to generate content, providing businesses with a wealth of opportunities to generate content, streamline their processes, reduce costs and enhance their operations.
Andrew Ng, a renowned computer scientist and Coursera founder, said that ‘Artificial intelligence is the new electricity’, and that not a single industry will escape the AI transformation that will take effect within the next few years.
Just as electricity and the birth of the internet transformed the world in past generations, Generative AI is poised to do the same. Therefore, making the best decision possible when it comes to incorporating Generative AI into businesses is crucial for staying ahead of the curve and remaining competitive.
But what could be achieved for companies who not only embrace Generative AI into their businesses but also go one step further by utilising it to drive real-world solutions? Enter Future Processing.
Future Processing – Generative AI products
At Future Processing, we use a number of Generative AI products. One that particularly stands out is Amazon Bedrock, an innovative tool offered by Amazon Web Services which offers an easy-to-use developer experience to work with a broad range of high-performing foundation models, like Claude 2 from Anthropic.
Using AWS and Claude 2, we are able to deliver artificial intelligence and machine learning solutions for businesses by providing expertise in predictive maintenance, recommendation engines, image analysis, NLP chatbot and sentiment analysis, and timeseries. We look at businesses’ case designs, work on data access and engineering, deliver a suitable model, monitor and manage the model, and then finally develop the full application.
By integrating the foundation model Claude 2 into our GenAI toolkit, we are able to create interactive PoCs of companies’ AI or ML products that verify the initial ideas. This allows us to make excellent, data-based decisions on our clients’ investments, before finally delivering a world-class product.
These products serve as the backbone for our transformative GenAI solution, which enables us to tackle complex data challenges head-on.
Why we turned to Generative AI
The Challenge
Our journey into Generative AI was driven by the needs of our clients. One of them had the issue of having to do lots of manual labour each year in order to feed his job search platform with data from their partners. This was a time-consuming, costly and inefficient process, so they asked us to create a digital product to automate this process. Traditionally, programming for this type of task would have been far too expensive (if possible at all). However, Generative AI opened the door to discovering a better solution.
It was for a good reason that we searched for a better solution to this issue – typically, our clients would need to create a new adaptor for these partner websites and adapt them in the case of any changes (e.g. structures, formatting, updates, and so on). This was time-consuming, costly and not an efficient way to operate. In order to avoid this issue, we came up with a general solution that would create the same output for every website from a variety of sources.
Traditional programming had always been a barrier to creating such a product due to the costs involved with building and maintaining an algorithm that would be required for such a versatile and changing environment. However, now with the aid of the more general capabilities of Generative AI, we have finally been able to bring this creation to life.
The Solution: Leveraging GenAI and Amazon Bedrock
Thanks to the help of Generative AI, we were able to build a solution that proves it is possible to solve this issue in a straightforward and efficient manner.
We created a simple web application that allows users to input the URL of a job offer they want to migrate to the new partner system. After pasting the direct web link, we simply press the “Process” button to begin.
Once processed, the AI scrapes the webpage in the background, using Amazon Bedrock Model to extract the necessary information. The output is presented in a structured JSON format that is retrieved from the screen.
A new window then opens with the completed form already filled out. This is done through the use of an RPA solution (Robotic Process Automation).
By using this model, we were able to present a successful proof of concept that demonstrates that the desired state can be delivered in just a few days – much faster than anything that could have been achieved without the use of Generative AI. This model has been completed and is now ready to be fully rolled out for use in production. The best part? It took just a few weeks of engineering work!
Expected Outcomes and Impact
Our GenAI solution has achieved remarkable results. It has not only automated expensive tasks that were previously performed by individuals. This has significantly streamlined the data-gathering processes for our clients.
By leveraging Generative AI tools, we’ve demonstrated that data aggregation from various sources can be standardised, saving time, reducing operational costs and boosting efficiency.
What our experts say
Krzysztof Szabelski, Head of Technology at Future Processing, sums our paradigm-shifting solution up very well:
‘We are building a solution with broad range of high-performing foundation models from leading AI companies available through Amazon Bedrock that saves our client $80,000 USD a year. GenAI helped us prove to our partner that tasks done by a large number of individuals worth $80,000 USD a year can be automated in just a couple of weeks of engineering work’.Krzysztof SzabelskiHead of Technology at Future Processing
Future Processing – our story
Founded with a mission to push the boundaries of technology, Future Processing has been at the forefront of innovation for over two decades. As cutting-edge technology consultants, software engineers and delivery partners, we are at the very forefront of this AI revolution and well-poised to assist businesses with their AI and ML needs. Our story is one of constant evolution, where we’ve harnessed the power of Generative AI to create game-changing products that save our clients both time and money.
Our journey has been marked by a relentless pursuit of excellence, and today we stand as a trusted technology partner who has successfully delivered business solutions to numerous companies around the world. With a strong commitment to delivering tangible business value, we’ve consistently pushed the envelope, driving innovation through technology.
Looking to the future
As we continue to refine our GenAI solution, the future is bright at Future Processing. Our huge success in the case study herein illustrates the transformative power of Generative AI. It underscores how Generative AI, backed by AWS, Amazon Bedrock and Claude 2, is poised to redefine the way businesses handle data. It’s a reminder that innovation isn’t a one-time event but a continuous journey. Embracing technology and adapting it to meet real-world needs is the key to staying ahead in the digital age.
We are firmly committed to exploring new possibilities and innovative applications of this technology (and many others), all while delivering exceptional value to our clients and the industry as a whole.
To learn more about our GenAI solution and our AWS partnership, visit other articles on our blog, explore our AI and ML services, and get in touch to discover the power of our AWS-powered cloud services.