Interview: How Imperative Chemical Partners Is Applying AI to Optimize Field Operations

AI has the potential to fundamentally change how oil and gas organizations operate. But, only if the data foundations are right. 

In this interview, Raymond Mitten, Vice President of Advanced Digital Technologies at Imperative Chemical Partners, explains how the organization is using generative AI, machine learning and secure data ecosystems to move beyond traditional reporting and enable faster, more informed decisions across field operations.

Lily Mae Pacey: Can you tell me a little bit about your role at Imperative Chemical Partners and share how it connects to the company’s broader AI goals?      

Raymond Mitten, Vice President, Advanced Digital Technologies - Business Transformation: I serve as the Vice President of Advanced Digital Technologies at Imperative Chemical Partners, where I lead the entire digital organization. My responsibilities span innovation, digital transformation and the full spectrum of operational technology support.

When it comes to AI, my focus is on building the roadmaps and platforms that support our largest and most critical teams, which is field operations. Our field teams are the foundation of our business, so my AI strategy is focused on enhancing the tools, data and systems that empower them to work more safely, efficiently and effectively. 
 
Lily Mae Pacey: What are the biggest challenges in building a shared data system when different operators use different standards?  

Raymond Mitten, Vice President, Advanced Digital Technologies - Business Transformation: It's a great question. Even within a large operator, even a supermajor, different teams often use completely different terminology. A field team in the Permian Basin might refer to a well asset one way, while teams in the Bakken or the Eagle Ford use entirely different nomenclature, all within the same organization.

"Operators who choose not to participate in the data‑sharing and AI innovation wave are the ones most likely to fall behind."

That variation makes data alignment a complex challenge. Our approach is to begin with industry standards, regardless of how the data comes in or what someone calls something locally. We normalize and map everything back to recognized oil and gas industry standard data formats. This ensures consistency, comparability and reliability across all our platforms and workflows. 

Lily Mae Pacey: How do you keep data secure while still making it useful and connected for AI-driven insights? 

Raymond Mitten, Vice President, Advanced Digital Technologies - Business Transformation: We primarily rely on Microsoft Azure’s industry-standard security frameworks and foundational security services. In addition, we’ve built architecture that enforces strict data isolation for any customer-specific, critical data. When there’s a need to combine a dataset, we obscure identifying details and focus on the aggregated outcomes, the analytics and insights, rather than on any named operator. In those cases, analysis is done at a regional or area level where similar characteristics exist. Overall, maintaining strong data isolation between customers is a key principle of our approach. 

Lily Mae Pacey: What role does a cloud-based data lake play in creating scalable and secure data ecosystems? Especially in multi-operator environments. 

Raymond Mitten, Vice President, Advanced Digital Technologies - Business Transformation: It’s no secret that the real power of the data lake isn’t simply having data from different operators, it’s the ability to bring together many different types of data. For example, we collect GPS driving data from our trucks as they visit various operators and locations. That information becomes less about any specific operator and more about understanding logistics and optimizing chemical management programs.

The data lake also lets us correlate operational activities. For example, take water sampling: our field teams collect samples, those go to the lab and we can begin identifying patterns in chemical usage based on those results. So, while the data does come from multiple operators and sources, the value is in the broader collection, the patterns and correlations that emerge when all this information is analyzed together. You simply wouldn’t discover those insights if the data were siloed in a single application or tied only to a single operator. 

Building standardized, scalable data foundations is increasingly seen as a prerequisite for intelligent asset management. These themes will be explored in more depth at the Intelligent Asset Management in Energy Summit!

Lily Mae Pacey: How important are partnerships in creating a robust data-sharing framework? What three key qualities do you look for in these collaborations? 

Raymond Mitten, Vice President, Advanced Digital Technologies - Business Transformation: Strong data partnerships are essential not only for the work we do, but also for operators who rely on those partnerships to gain insights they wouldn’t otherwise have access to. It’s mutually beneficial.  

For any partnership to work, the operator must clearly understand the value: what problem they’re trying to solve, what outcomes they want and why the insight matters to them. Their level of technical and cultural maturity also plays a role. For example, some supermajors are extremely advanced technically. They have analytics teams, API integrations and internal development capabilities. However, they may lack maturity in a specific subject-matter domain, which limits their ability to see the value in forming a data partnership.

On the other end of the spectrum, you might have a smaller operator who simply doesn’t believe the gains justify sharing their data, those who haven’t seen enough margin impact. In those cases, I’d argue they don’t yet understand the full potential. Then, there are the smaller operators working under much tighter margins, even a 2% improvement in efficiency could be the difference between continuing operations or shutting down. For them, the incentive to pursue insights is often much stronger. 

The reality is that the value is there for both large and small operators. It just depends on what they’re trying to achieve and whether they can recognize the patterns and insights that the data can reveal. 

Lily Mae Pacey: How are you using both generative AI and traditional machine‑learning techniques, such as neural networks, to translate diverse datasets into industry‑standard formats, as well as to uncover operational patterns in areas like chemical usage, production volumes and automated injection? 

Raymond Mitten, Vice President, Advanced Digital Technologies - Business Transformation: We use generative AI for specific purposes. One major application is data translation, taking a partner’s dataset and mapping it to industry-standard patterns. As operators have different data structures and data flows, it can take an individual a long time to manually work through all those mappings. Generative AI helps automate that process and align the data with an industry-standard oil and gas data model. 

We also use generative AI to create new outputs. For example, it can generate an optimized routing pattern for a driver based on chemical usage, delivery schedules, timing and other optional factors. That’s really where generative AI excels, creating something new. 

When we’re identifying patterns in production volumes, chemical usage, automated injection and similar operational activities, that relies less on generative AI and more on traditional machine learning and neural networks. These techniques are designed to detect patterns and correlations across data, relationships that no human could manually identify.

What’s interesting is that the same underlying technologies that power generative AI also support machine-learning models we use. However, in this case, we’re applying those foundational technologies specifically to pattern recognition and correlation analysis, rather than to generating new content. 

Lily Mae Pacey: What are the benefits of using AI to create a unified view of asset behaviour across operators? How does this impact decision-making? 

Raymond Mitten, Vice President, Advanced Digital Technologies - Business Transformation: We collect a tremendous amount of data in oil and gas and traditionally we use that data to produce reports. People read those reports, interpret them and then decide what actions to take. In reality, reports are just an interim step, a temporary mechanism to help us understand the data before making a decision. 

AI changes that. It helps us move directly from data to action. In particular, generative AI can take raw data, add content and effectively bypass the traditional reporting cycle. Instead of teams spending months reviewing reports, analyzing trends and determining what to do next, AI can surface those recommended actions immediately.

Ultimately, this isn't unique to oil and gas. The same applies across the energy grid, field operations and manufacturing. The value isn’t in the report, it’s in the action taken. The faster an organization can get from data to action, the faster it can achieve real optimization. AI allows us to skip several of the manual, time-consuming steps that used to stand in the way of that. 

Lily Mae Pacey: How can regional modeling unlock in efficiencies that individual operators might miss when working alone? 

Raymond Mitten, Vice President, Advanced Digital Technologies - Business Transformation: Absolutely. A real-world example I often use is when you have two assets sitting almost side by side, operated by different companies. On each site, a field worker is ultimately responsible for managing the well and that person may have 25 or 30 years of experience. They rely on the chemical program and mixtures they’ve always used, simply because that’s what has worked for them historically.

But scientifically, those two assets, due to their proximity, often have very similar geology, water quality and other conditions at that point in time. Despite this, the operators may still use completely different chemicals, mixtures and volumes purely due to habit or legacy practices.

Operational excellence in oil and gas is increasingly driven by data, analytics and cross‑asset visibility, all key themes at the Operational Excellence in Oil and Gas Summit!

What advanced analytics allows us to do is strip away those preferences and focus on science. By analyzing water samples, production volumes and other technical signals, we can determine which chemical program aligns with the underlying geology and conditions. It’s not about the individual decision-maker, it’s about identifying the correlations that drive higher production based on what’s truly happening in the asset. 

Lily Mae Pacey: In your opinion, what does the journey for AI-driven data integration in multi-operator environments look like over the next five years? 

Raymond Mitten, Vice President, Advanced Digital Technologies - Business Transformation: To reach the level of insight companies are aiming for, there has to be a strong data-sharing partnership model. To be clear, this isn’t about operators sharing data with each other. It’s about operators sharing data with the service providers they rely on because that’s where real optimization happens.

When I compare operators who share near real-time data with their service partners to those who don’t, the production gains are significant. I won’t cite specific numbers, but improvements commonly range from about 2% up to 30%, depending on the asset. Even a 2% lift can be huge from a revenue standpoint. On assets that lack modern technology, like automated injection or tank‑level monitoring, the production gains can be at the higher end of that range. 

"Even a 2% improvement can be the difference between continuing operations or shutting down."

Some of that improvement comes from the fact that those assets weren’t optimized to begin with, but the real driver is visibility. When data is shared, acted in real-time and tied to accurate schedules, volumes and chemistry, the production increases can be substantial.

Lily Mae Pacey: Do you think that level of integration is achievable within the next five years? 

Raymond Mitten, Vice President, Advanced Digital Technologies - Business Transformation: I see significant momentum, especially among the larger operators, toward data standardization and data-sharing platforms. I’m not sure we’ll see full adoption across the entire oil and gas industry in five years, but the direction is clear. Operators who choose not to participate in that innovation wave, or who delay, are the ones most likely to fall behind. 

Lily Mae Pacey: You recently participated in the AI in Energy Summit. What prompted you to get involved and why do you think forums like this are important for advancing AI adoption in oil and gas?

Raymond Mitten, Vice President, Advanced Digital Technologies - Business Transformation: I think it’s incredibly important for people in the oil and gas industry to come together, share ideas and hear what others are working on. You can research on your own, but you won’t get the same insights you gain from real conversations. Many experts have valuable knowledge they’ve never published and talking with them can spark ideas you can take back and apply. For me, these events also support one of my key initiatives, driving data partnership and sharing across the industry because that alone can create significant industry progress. 

Data standardization, secure data‑sharing and moving from AI experimentation to real operational impact will be further explored further at the AI in Energy Summit! Join us for the next edition of our event in Houston.