Moacyr Galo, head of labs, outlines the AI journey for law firms, moving from data to wisdom, using machine learning to make the data connections and the decision-making properties of knowledge graphs.

Artificial Intelligence is a new set of functionalities that will push our technology-based decision-making process to a predictive future. Taking the right approach, AI can be a solution that brings value to executives in complex waters with easy-to-use tools regardless of where you work.

At, we have extensive experience led by Moacyr Galo, head of labs, moving from the standard maturity model and the journey from data to wisdom. In this Q&A, Moacyr describes the decision-making properties of knowledge graphs.

Q: What does the next decade look like for Artificial Intelligence?

A: In the next 10 years, the leaders among each segment will play sophisticated, highly competitive games using very sophisticated algorithms. The industry is digitizing everything in real time. Thought leaders of those corporate clients will demand new service providers who will need to provide new types of services.

Take legal services for example. There will be smart contracts where the execution of the agreements will be done automatically. So, understanding how to bring value to this real-time world will be key for the lawyers of the future.

In all industries, we will compete to provide the services that will be demanded by companies. This will be basic, and we need to prepare for them. If we don’t start collating data now and getting in the middle of that change, we cannot use the tools of today to get there. We must understand what the scenarios and research are and try to adapt those tools and you must start the journey now.

AI will not replace people – that is a fallacy – AI will just empower you and in labs, we are building a framework called AIR (Augmented Intelligence Research) for that challenge.

Q: What is the difference between AI versus BI?

A: BI is basically a set of tools that looks at today’s data and the past data to be presented to you in dashboards or reports. We call it “descriptive analytics”. AI is “predictive” – it looks at that same data and combines it with additional internal and external business contextual data, to make predictions and decisions for the future.

Q: What is empower Decision Intelligence?

A: empower Decision Intelligence is a sophisticated cloud-based tool from labs that predicts future scenarios across business functions connecting CRM to Finance and the variables within, such as cashflows and profitability. Built specifically for the Professional Services industry, the scenarios can be based not just on past transactional data but also other internal/external based simulation inputs. This allows for improved and more real-world prediction scenarios to empower executives (CEOs, CFOs, and COOs) to strategize better for the future.

Q: You have a journey map presented below. How do you describe each step in the “Journey from Decision to Outcome”?

A: All companies need to turn the data and knowledge into an asset. This is an image of how we are going to solve that problem. Let me take you through the journey:

Data: I will explain this using a real client that is in a good maturity technology stage. They are doing everything they can to leverage their data in a regular cadence using external tools and developing internally to develop their dashboards. The problem is what they have is very difficult. To run their business, they have several systems. For the most part, companies have a lot of legacy systems internally including external vendor systems. Generally, those systems are not integrated, and you miss opportunities and invest a lot of time to get them to talk to each other. They must go to spreadsheets and that is when the knowledge you have in the data cannot be fully utilized. The knowledge is there but you cannot turn it into something useful. You must mine for it generating spreadsheets. Users are looking for a miracle tool that can do that for them because they realize they are at the limit and are looking for a way to get meaningful data from those isolated silos.

Information: Here, companies are at the limit of extracting value from the data using the tools that they have today because the premise is to present the dashboards or the reports that users must go into to get the data. But like drilling for oil, it is too deep. So, they start fracking. They must get that asset and to extract that hidden asset costs huge amount of money. They know it is there, they see the dots, but they see them in different colors which tells them the data is not connected.

Knowledge:What we are doing here is using algorithms to connect the data that today is disconnected, by using machine learning and a knowledge graph. The idea of the knowledge graph is to map the relationships between entities and to use algorithms to connect them. To transform information into knowledge, you must look at the whole context of the business and then when we connect, we have the knowledge available. To do this requires specific models’ entities broken into structured and non-structured data, and stripping the data using SQL data or a spreadsheet, mostly text. The most part of the data is stripping the structured data asking how many views, how many emails, how much more valuable informative data is there. We have much more knowledge on unstructured data and the idea is to use machine learning to make the connections to those. That is the value of the knowledge graph – to map the entities where you have the relationships of those entities, to look at isolated silos, and connect that data.

Insight: You will discover a lot of insights that will feel like a new world. You see we have a connection here, and I found another connection there. Then you start learning from that connection. You open new possibilities to detect the correlations between those previously disconnected silos. Machine learning will perform a deep conceptual understanding here.

Wisdom: At the end, as more understanding correlates, you are at a stage of exploring AI where we have three different solutions:

  • “Descriptive Analytics” (Power BI) just showing what happened in the past. It is a mirror of the past.
  • When we start using machine learning for “Predictive Analytics” is where we can predict what is going to happen. We learn from the past and try to predict what is going to happen next.
  • After a while of using “Prescriptive Analytics”, a model can recommend the next best action for you.

These are the three phases – the descriptive and dashboards and reports. Predictive is the same KPIs as you have had in the past but instead of looking into past, we look at the future where with prescriptive we can go to the model of what is suggested for you.