My Takeaways from Big Data LDN 2024

A week of thought validation and mature AI thinking!

The 2024 Big Data LDN conference was the second year that our EncompaaS team have had the opportunity and pleasure to exhibit, and it was plain to see the speed of change in the market since last year’s event.

In 2023, the GenAI boom was bursting into life and the buzz around the event focused on how to take advantage of such powerful technology and where AI could be used to deliver tangible results against investment, something key stakeholders are monitoring. My position following the event was clear – data quality and governance were going to play a significant role in the success of AI and at that point, we went on a journey at EncompaaS to help our clients prepare their information.

This year, there was a shift in focus to a data quality driven approach to AI and an emphasis on compliant adoption. A small change but a change nonetheless, aimed at preparing and understanding data, a position we’ve taken at EncompaaS since GenAI’s inception into the market.

But why this pivot in thinking in 2024 to an increasingly mature AI focus? Gartner estimates that at least one third of GenAI projects “will be abandoned after proof of concept by the end of 2025, due to poor data quality, inadequate risk controls, escalating costs or unclear business value projects”. I believe this figure could be higher and recently posted on the topic here. The reasoning behind this is that in many cases POC environments are like chalk and cheese in comparison to live, production environments and thus output from those POCs will not reflect real-life processes and business value.

The challenge for most companies is that they know they have a huge amount of data and they want to maximise the impact of AI, however they don’t know what they have and where, as well as its context and business value. Without knowing the answers to these questions, it will be challenging to see tangible success from AI, as costs will spiral, outcomes will be substandard and importantly, responses could be false.

To keep up with market and business demands around AI, data quality and governance must be a fundamental part of an organisation’s AI posture and AI Acts across the globe should support this thinking moving forward. But how did this reflect in the output from Big Data LDN 2024? Well, here are my three key takeaways:

1. AI use cases are going through iterations and focusing on specificity: Use cases around entity extraction are becoming a common thread. I had many discussions at the event around how information can be curated and prepared to ensure the right information is being used at the right time. In a legal contracts use case, it will become imperative to understand where contracts are located, as well as contextual and value-based understanding to feed to AI tools. Costs will increase if datasets are large and less accurate due to lack of visibility and the burden placed on AI recognising this. It was great to see information curation as a popular topic.
   
2. Retrieval Augmented Generation (RAG) is becoming increasingly important as AI specificity grows: Out of the box LLMs are often not specific enough for an organisation’s business needs, lacking internal business knowledge. This can be challenging when implementing internal chatbots, for example. RAG was a hot subject of conversation at Big Data LDN, alongside data preparation. Many of the delegates I spoke to recognised the importance of source information being prepared correctly to enhance the use of AI in their business and acknowledged RAG would be playing a significant role for them moving forward to drive accurate outcomes.
   
3. Creating strong data foundations and governance principals remains the key to success: My gut feeling and thoughts from 2023 are validated. Many companies are keen to adopt new AI projects but their foundational data knowledge and visibility aren’t strong enough and many voiced a concern of “I don’t know what I have and where” as the core compliance and productivity issue stalling their AI progress. The end goal will be different in every business, however the foundations need to remain the same – understanding your information will provide the basis you need to succeed with AI, whatever success looks like.

Data preparation is one of our core capabilities at EncompaaS so I can encourage you to reach out to me if you’d like to discuss how we can help your organisation generate data quality and de-risk your information in readiness for AI.

I’m already looking forward to next year’s event and seeing how the market continues to pivot, mature and address these important AI topics. Finally, collaboration is key in our space, as technologies intertwine to deliver value to customers – having built some excellent, new relationships, I am excited to explore what will come of these in the year ahead!