Neeraj Koul, Chief Data Officer at the Federal Railroad Administration (FRA), speaks with Ranyah Salous, Director of Digital Solutions, Data and AI at Guidehouse, in a video interview about his role, the importance of reliable data, measures taken to ensure effective data scaling, having a multi-faceted and aligned approach, AI’s role in FRA, and the need for proper governance and training.

Koul serves as a Senior Technical Advisor at the Federal Railroad Administration and also holds the role of CDO. With a PhD in computer science, he has nearly two decades of experience, and his focus has been on harnessing data to drive mission success. He has worked across academia, the private sector, non-profit organizations, and the federal government.

As FRA’s CDO, Koul provides strategic guidance on all data-related initiatives within the agency.

When asked how reliable and timely data contribute to advancing FRA’s mission, he notes that FRA’s mission is to ensure safe, reliable, and efficient movement of people and goods, strengthening the present and future of America.

Adding on, Koul states that data plays a crucial role in supporting decision-makers, inspectors, and grant teams in achieving this mission. Elaborating, he states that FRA has been collecting safety data from railroads for a long time. Two particularly important datasets it gathers include information on railroad incidents and accidents.

Whenever an incident occurs that meets a specific criterion, railroads are required to report the details to FRA. Similarly, any casualty, whether an injury or fatality, within certain parameters, must also be reported. FRA utilizes this data in multiple ways, primarily to identify trends, assess safety metrics, and uncover potential risks.

Emphasizing the measures FRA is taking to ensure data scales effectively, he mentions three approaches.

1. Infrastructure: It ensures that the administration has the necessary computing power and technological framework to manage, process, and analyze the data that comes in.

   This is essentially the IT component — making sure that systems are capable of handling large volumes of data efficiently. To support this, FRA has adopted best practices like a cloud-first architecture to enhance scalability, reliability, and performance.

2. Data management: FRA recognizes that IT infrastructure and data management are distinct but equally critical areas. Data management ensures that data is high-quality, reliable, and properly governed. This involves establishing policies, procedures, and best practices to maintain data integrity as it flows through FRA, enabling accurate and effective analysis.

3. Analytics and decision-making: With a robust infrastructure and well-managed data, the next step is ensuring that insights reach the right people at the right time. Different stakeholders require different levels of detail. For example, a Chief Safety Officer may need high-level trend analysis, whereas a data scientist or operations researcher may require more granular, technical data.

FRA ensures that data is accessible through visualization tools and tailored analytics, allowing decision-makers and subject matter experts to extract meaningful insights that drive their daily operations.

Having a multi-faceted approach is the key, says Koul, and data is a means to achieve mission objectives; it is a tool. The biggest challenges arise when the key components — infrastructure, data management, and analytics — are not aligned.

Any gap or weakness in one of these areas can significantly hinder the ability to extract value from the data. However, when these elements work together seamlessly, the volume and complexity of data can be effectively managed, states Koul. He notes that while FRA does not have streaming data as of yet, it is harnessing what is traditionally called a variety of data to meet fine insights.

Sharing his opinion on how AI enhances FRA’s ability to make data-driven decisions, Koul stresses that AI has gone through multiple waves over the past 60 years. He continues that its potential is truly being realized with the advancements in Generative AI. The commoditization of AI has enabled new users to leverage its capabilities.

At FRA, AI is a powerful tool that can help achieve mission objectives, Koul reiterates. To truly harness its potential, it must be integrated into daily operations while also managing the risks and challenges associated with its use.

Koul explains AI’s role at FRA in two key areas:

1. Internal applications: Using AI within FRA and the federal government to enhance day-to-day operations.

2. Regulatory oversight: As a regulatory agency, FRA must also ensure the safe and responsible integration of AI-driven tools in the rail industry.

One of AI’s greatest promises is its ability to reduce manual workload and increase efficiency. By automating repetitive tasks, AI allows employees to focus on higher-value work. Apart from productivity, it is a must to ensure that it is used the right way and no data is leaked. For this purpose, Koul reveals that the administration is collaborating with the Department of Transportation to develop AI tools that subject matter experts can use in a ChatGPT or Gemini-like setting.

FRA is making sure that the available tools are safe to use at an enterprise level. Koul envisions human-machine teaming, where AI and subject matter experts collaborate for more informed decision-making.

Wrapping up, Koul states that with AI tools becoming more accessible to end users, the barrier to entry has significantly lowered. This makes it even more important to have proper governance and training in place. Users need to understand how to effectively use these tools, ensure that the insights they generate are appropriate for their specific use cases, and account for potential challenges and limitations.

Concluding, Koul states that as a CDO, his role is to balance these aspects of promoting AI adoption while ensuring responsible and effective usage.

CDO Magazine appreciates Neeraj Koul for sharing his insights with our global community.