Reworld™

Cal Link

AI Governance Builder

Cal Link joined Reworld™ in 2019 to lead Data and Analytics, specializing in applied AI and enterprise data ecosystems. With 30+ years of leadership across Marketing, Finance, HR, and IT, he previously served as Chief Data Officer at AIG Retirement Services, driving data strategy, analytics, governance and platforms. He is a frequent speaker on AI and data supply chains.

The old computer programming saying "garbage in, garbage out," which was popularized back in the 1950s and 1960s, remains relevant today in the rapidly advancing world of artificial intelligence. As organizations rush to implement advanced analytics and AI, they sometimes overlook the fact that without integrated, reliable data, results may fall short of delivering meaningful impact. In other words, the insights and effectiveness of AI are only going to be as good as the data we provide it.

At Reworld™, we are a leader in sustainable waste solutions and zero-waste-to-landfill programs. To achieve our goals, we reimagine waste and how to approach our clients' unique waste streams. We took this same custom approach to our data and analytics. We focused on building an enterprise-wide data foundation that would support faster, more informed decision-making across the organization.

A cornerstone of this foundation was the development of what we call the Reworld™ Data Hub, a centralized data system that connects information across the enterprise. To do this, we collected and organized the most relevant data, then tackled the more complex challenge of integrating it across our systems and teams. But we didn't stop once this was done; by strengthening this foundation over time, we have made advanced analytics and AI more accessible and actionable across the business, from our corporate headquarters in New Jersey to our network of more than 90 facilities across the country.

Enterprise-Wide Data Integration

The Reworld™ Data Hub integrates operational, financial, commercial, and facility data across the enterprise. This program drives the connectivity of all company data, serving as the backbone for our analytics and AI. My team built the Data Hub gradually, carefully increasing its depth to support faster insights, which, in turn, expedite informed decision-making from months and weeks to days and hours. We leveraged Qlik data solutions and Snowflake as the core technologies. Shared, reliable data has improved alignment and execution across teams, supporting almost every function of the business. However, this was only achievable by developing the hub one project at a time, knowing that AI can only add purposeful value if it’s grounded by trustworthy data.

 As organizations rush to implement advanced analytics and AI, they sometimes overlook the fact that without integrated, reliable data, results may fall short of delivering meaningful impact. 

From Data to Impact: Operational Results

Data can drive measurable improvements across a business in many ways. While the applications of data are limitless, at Reworld™, several notable use cases demonstrate its value in making us safer, more efficient, and more effective overall at improving sustainability outcomes for our customers:

• Boiler Efficiency and Reducing Downtime: Research data enabled our facilities to reduce boiler cleanings from four to three times a year. AI-enabled analytics showed we could reduce downtime and labor costs without sacrificing the equipment's performance and reliability.

• Dioxin Emissions Modeling: Enterprise data is helping us identify the causal factors of dioxin emissions. With new EPA standards lowering thresholds, the Reworld™ Data Hub is enabling us to proactively develop models that prevent potential emissions risks.

• Metals Price Forecasting: We are developing predictive models to help our sales team forecast the pricing of metals that we recover from Reworld™ Thermomechanical Treatment Facilities (TTFs). This is equipping the team with more effective hedging strategies for the metals we recover and sell. So far, these predictions are accurate within two or three percentage points looking six months ahead.

• Fire Threat Identification: To determine fire risks across our facilities, we built a fire threat model using unstructured data from operational reports and logs. This provided us with insight into potential fire hazards and recommendations for measures to help prevent incidents.

Governance as “Data Supply Chain Management”

At Reworld™, we view the governance of data like the management of a supply chain, ensuring that data flows reliably from where it is created to where it is used, with clear ownership and controls throughout the process. Governance is focused on outcomes, which allows us to assess data quality based on how it is (or will be) used. Both governance and AI must work together, since AI can only add reliable value when it uses clean, reliable data. A lack of governance would feed unsecured data to AI models, leading to inaccurate information and misinformed decisions. As AI continues to evolve, governance will need to remain stringent.

AI as a Tool, Not a Replacement

As useful as AI is, it necessitates strict human guidance and oversight. AI should be seen as an intern who needs clear direction, not an executive with a trusted vision. This approach prevents misinformed decisions by ensuring AI users think critically about the data they use and the outputs of our machines. In practice, this is how Reworld™ maintains a clean, trustworthy database and prevents unsupervised use. AI is only a value-add when data quality is high and business objectives are clear. Then the real differentiator is not the tool itself but how users implement it.

The Future of Data and AI Integration

Properly constructed and engineered enterprise data programs enable faster, more informed decision-making across every level of an organization, from daily operations to long-term strategy. At Reworld™, our strong foundational data is helping us improve performance across our network while supporting our customers’ sustainability goals.

In many ways, the challenges companies face when looking at their data are similar to those in their waste streams; without the right systems in place, valuable resources are lost. With the right foundation, they can be recovered, refined, and put to far better use. As AI capabilities continue to evolve, success will depend on how well organizations integrate and govern their data. Leaders must also approach new use cases with a clear understanding of their objectives and a disciplined, cautious mindset to ensure outputs are reliable.