The Center for Data Innovation spoke with Rio's CTO Brian Flynn about data technology and AI and reporting
Written by Rio ESG
The Center for Data Innovation spoke with Brian Flynn, Chief Technology Officer for Rio ESG. Rio ESG is a sustainability-focused software platform that combines software, consultancy, education tools, and other features to improve their clients’ sustainability performance. Rio helps various industries including investment managers, corporates, financial services and public sector organisations track sustainability data, report on ESG, manage policies, and learn how to become more sustainable.
Kir Nuthi: Could you describe Rio and how Rio uses AI and reporting to apply its carbon footprint and sustainability data?
Brian Flynn: We are a sustainability consultancy with a platform called Rio, which we use as part of our hybrid approach to providing sustainability services to organisations. As the CTO, I focus on developing the platform side of things so that it helps augment the services that the consultants within the company can offer. We also have clients who interact directly with the platform, so ensuring their experiences are good is critical.
Rio is a pretty comprehensive, extensive platform. It is a mechanism for organisations to improve their sustainable practices by assessing their current position, understanding that, and then using the system to help move forward and perform better. We have engineered the product so that data is central to everything we do. The system has several components or parts, but it always comes back to the data. We undergo quite an extensive onboarding period to get the data into the system and ensure the data is of the right quality. Our manual consultants work with the clients to help them understand the context, what kind of data we’re looking to bring in, and the reasoning, so our clients have a broader perspective on what goal to aim for regarding sustainability. The Engage component of our system is a learning management system within Rio. Engage helps Chief Sustainability Officers and people whose day-to-day sustainability requirements mean a more comprehensive understanding of which regulatory aspects and frameworks will be most relevant to their goal. Also, it’s a platform for helping educate the wider workforce and getting buy-in across the whole organisation.
Once we’ve got to the point where we have the data ready to be ingested into the system, we have two approaches: connect via APIs to other systems where interoperability is vital or give clients the option of uploading it directly into our system. Through templates, tutorials, and good user design, we make it as pain-free as possible to bring better data quality and data into the system. After this is built up, we report on the collected data. We don’t just bring in business intelligence but also report against frameworks important to investors and the governance structures that affect them. By doing so, we’re helping organisations meet their legal requirements and match the reports and expectations of their investors.
The AI we predominantly use within Rio is not the kind of AI that’s entirely reliant on data. There are two main types of AI to fill families within AI. One is statistical AI, which comes under the umbrella of neural networks, and deep learning, but needs big datasets to get anything useful out of it. We don’t feel there’s huge value within that for us beyond anomaly detection within transactional data coming into the system and perhaps filling in gaps in data. But there are issues about data integrity when you have certain levels of data that are artificially created SAS amongst your natural data. You can lose the ability to claim that the data quality is above a certain level.
Our preferred approach is to use AI in a rules-based approach. It’s about capturing knowledge, capturing expertise, and mapping against the rules. And once we’ve created these, these rules-based models automate processes within the platform. So an example of how we’re doing that is our materiality assessments. At the early stage of a consultancy, we can use AI to understand what’s material, what’s important, and what’s relevant and to report against that for an organisation. We can work with our clients through this rules-based model we’ve created by eliciting knowledge using our proprietary methodology.
In the end, the organisation gets tailored reports with actionable insights to discuss with a Rio ESG consultant and then decides to take the process further with a real basis of information. This is something we’re looking to build upon. The beauty of the technology is that it’s scalable. So as long as we can keep fitting in that expertise, we can keep building out these models and capturing areas of consultation that are ripe for automation.
Nuthi: What makes you different from other ESGs on the market? What’s Rio’s unique approach that’s different from other competitors?
Flynn: One way is the breadth of what the system does, so when eventually engaged, we can help an organisation understand what it’s doing from a sustainability standpoint. From what the board of directors has committed to on behalf of the organisation to how they face global challenges, we have a whole section of our consultancy dedicated to our clients gaining that kind of level of understanding.
Through our data-focused reporting, we have high-quality dashboards that our clients expect. But that’s a given, almost. We go beyond our competitors with the granularity of the data we capture. We have a massive amount of transactional data for each of our customers. So we can feed in data from all sorts of sources. We connect to utility meters, gas meters, electricity meters, and more to work in reporting against these primary sources at sometimes even half-hour intervals. So behind this is a huge data warehouse, which is set up to handle billions of rows of data.
The advantage for our clients from us capturing all that data is that they can gain greater insights into an organisation’s position that you wouldn’t necessarily have if they were reporting at a very basic level. You find these hidden realities through this discovery process, which you don’t necessarily get by having high-level overview data. It means that we’re able to kind of track progress more quickly as well. And where we’re connecting to external systems where we have that interoperability, once we’ve got a client setup, the system looks after itself and constantly feeds data. Our various automated, rules-based, and proactive systems look for patterns within the data, so we can flush up alerts to make the client aware of things that need to be made aware.
We’re not entirely self-managing yet. But we’re moving towards a position where the combination of massive data that we have among the rules-based mechanisms can look after the client’s interests in a self-managing system. And our clients can have confidence in its design and strong ability to audit it and what those outputs are. We feel that, for example, if you’re overpaying for your waste service providers, we can explain down to a greater depth why that is and give them recommendations on what they do next.
Nuthi: What are some of the primary concerns and benefits you’ve encountered regarding using data in the UK regulatory environment?
Flynn: The challenges have generally been around the location of our data storage. We have to ensure that post-Brexit, our main data storage was UK based. While it was almost entirely, we still had to make sure everything was relocated into a central location. Regulation is not something we’ve faced very often, but compliance is more complicated for some companies. We started building up our client base when this was a beta product, and we were only working with smaller organisations, many SMEs and startups. That helped us refine the platform, enabled us to work with executives and less bureaucratic smaller organisations where everyone had access to the data and helped us understand the needs of that type of customer.
The next wave of clients were public sector organisations with very different ideas of what kind of system they wanted. Then, we began building towards a lot of what their requirements were. Public sector organisations have the strongest stipulations on the data side, so that data had to be located in the EU first and post-Brexit in the UK-based system we have today. Regarding GDPR, we were quite fortunate in that we don’t store personal data; if we do, it’s a fractional amount. This helps us sidestep a lot of the regulatory kind of considerations.
Beyond that, the other consideration we had was how data as an organisation is important to report back to the individual client who’s provided the data and to aggregate data across clients. That way, we can then find sector-specific insights and wider organisation-level insights. For us to do that, the more data we have, the better. So we had to ensure that if a client started to move on, we didn’t lose the benefit of having them as a client and having access to that data. We’ve had to look to anonymise data and find a process of anonymising the data so that there’s no danger of the client feeling that we’re holding on to data specific to them and identifiable back to them. This anonymisation ensures we can still use that data in constructive ways to build future models.
Nuthi: As you’re trying to expand, what are the challenges you’re facing in the expansion and outside of the UK markets?
Flynn: We’re now looking to work with finance companies, investment companies, and certain international companies. We’re still discovering what the challenges relating to data are internationally. Because of this, we’re no longer confined to working with UK-based organisations. We’re looking for a global reach into America, the East, and elsewhere. So this is currently a phase of discovery in terms of what the different regulatory environments are going to require of us. We have some ideas, but we’re still getting to the point where we fully understand what changes we need to make. I think one thing is that we won’t just have a single data storage location. We will have to look to service clients with data that is located close to there. That will be one of the first requirements as we look internationally.
Click here to read the full article on the website.