We recently caught up with Len Langsdorf, CTO of Digital Innovation Labs at CAPCO, to discuss the need for businesses to rethink one of their most important functions – how they make decisions – and how they can leverage AI to support better, faster decision-making.
According to Langsdorf, the average business is still using Microsoft Excel to store and analyze data. The same can be said for traditional analytics dashboards too. Business users prefer dashboards because they make data visual. 50% of our brains are dedicated to processing visual information and studies show that companies that use data visualization are leaders in revenue growth.
Despite their widespread popularity, dashboards are leaving business users with stale and untimely insights, error prone analysis and more questions than answers. The reason is, a dashboard can’t tell you why something happened, what decisions to make, or how to react to specific situations.
Decision intelligence is the answer to the legacy approach of creating spreadsheets and dashboards to access the data required to inform important day-to-day decisions. By analyzing cause and effect, decision intelligence transforms the massive amounts of data businesses collect today into insights + recommendations for action when and where stakeholders need them. With decision intelligence, you can go from insight to action 10x faster than with traditional BI tools.
What is Decision Intelligence
Decision intelligence is an evolving approach to analytics that provides business users with the data and recommendations for action they need, when they need it to support business decisions. “Engineered decision intelligence” is one of the top 10 trends in data analytics for 2021, according to Gartner. More than just analytics for occasional reporting, decision intelligence can support better outcomes the moment each decision is required.
Most of today’s decision-making processes tell you the “what” of your data — and not the “What do I do now?”
Decision Intelligence solves for decision latency, the time it takes to make a decision in response to a business change. It focuses on pulling clear, simple, and actionable insights from your data fabric — using artificial intelligence (AI) to parse through your data and recommend actions based on machine learning analytics and Contextual Intelligence. DI is capable of making immediate recommendations to support independent decisions while being able to fully explain and defend the recommendation with contextual insight.
Finance and AI solutions go hand-in-hand
“Finance and AI are almost completely intertwined”, explains Langsdorf and the finance industry has more to gain from Decision Intelligence than nearly any other sector.
First, financial organizations are packed with data. It’s not uncommon for a financial institution to be sitting on petabytes of data — and that’s a lot of potential insights to work from.
Second, this massive volume of data won’t be processed manually. Those millions of gigabytes are quite literally impossible for any data scientist or team of data analysts to ever parse through. This is where organizations will need to use decision intelligence along with advanced machine learning algorithms.
Third, the finance industry is built around numbers. This makes it a much easier industry for digitized solutions to tackle than the entertainment or tourist industries, for example (although it’s not impossible there either).
Fourthly and finally, much of the hard graft being done in the financial industry is work that would be far better suited to AI. There are many repetitive, menial tasks involving a lot of typing and looking over numbers. AI can take this work on itself, leaving the human staff to perform more high-value, meaningful work that plays to their strengths. Additionally, AI can help reduce biases humans run into when considering the types of decisions that need to be made.
Human and AI working together: The future of Decision Intelligence
According to Langsdorf, the biggest potential of Decision Intelligence solutions in finance is in how they can shift the role of industry professionals. Rather than doing operations, “people need to stop doing operations and start becoming auditors of operations”.
In other words, the AI should be doing the work. Leaving specialist talent to look at the results, answer questions, and help boost the AI’s performance.
This kind of collaboration isn’t unique to the finance sector, as it is generally believed that AI and machine learning won’t completely replace human workers, but rather supplement them and the work they’re already doing.
In the same way that using an app can make tasks easier, less error-prone, and produce better results, replacing traditional decision-making processes with AI and Decision Intelligence solutions can provide long-term benefits, especially when paired with an expert. As Langsdorf puts it “A human-in-the-loop AI system can help validate the decisions an AI system is making. That way, AI helps advance beyond operational decision-making to tactical decision-making. Both AI and humans can make bad decisions. Using both technology and humans together to connect the dots is where the biggest potential lies.”
NLP, AI, and the human perspective
So what will AI and human collaboration look like for the finance sector? Diwo has a solid grasp of this relationship.
AI is better suited to the data-heavy, mathematical work thanks to its ability to parse through vast pools of information without skipping a beat. It can also provide relevant context for that data with ease.
A human can then take those insights and act on them. Using Natural Language Processing (NLP), the human and Diwo’s AI can hold a conversation over the data, asking each other questions and learning from one another.
This is the missing piece of the decision-making puzzle that we referenced before: the “What now?” or “What next?”. Rather than a human analyzing the data and still having no idea how to act on it, Diwo’s Decision Intelligence platform analyzes the data and presents the best courses of action.
These routes are paired with recommendations, can be provided before a human would even know to consider those routes, and could include accurate predictions on how each route will play out.
Langsdorf describes this in perfect, simplified terms, as:
“If you’re using Diwo, you can ask simple questions and get answers using natural language. You can interact with the system, and it can learn from your inputs and changes. If you hire a data scientist and they create a report for you, there’s still a lot of back and forth to understand how and why the outcomes were reached. This produces a huge time lag between data and decision. Diwo eliminates that time-to-decision gap and helps you drill down into the rationale behind decisions without the need to interact with a data scientist.”