In order to embrace AI responsibly, teams need leaders who can create a sense of direction, establish guardrails and ensure there’s a balance between automation and human oversight.
For Joyce Ling, vice president and chief of staff to the CEO at home products manufacturer SharkNinja, guiding the company’s AI and machine learning team comes down to owning outcomes — and stepping back.
“I give people real autonomy to lead in the places where they’re most capable and most comfortable because that’s where they move fastest, and I stay close enough to be useful when it counts,” Ling said.
Meanwhile, at InterSystems, a company that develops data management, integration, and analytics technologies, leaders like Director of AI Enablement Nicholai Mitchko start every AI initiative by understanding the problem and the desired outcome before diving into development.
“We look for initiatives that can improve productivity, quality, customer outcomes or employee experience, then establish success measures before scaling them,” Mitchko said.
AI evolves quickly, and it can be challenging for teams to keep up. That’s why leaders at CodePath.org, a company that partners with universities and tech organizations to educate the next generation of AI-native technologists, ensure workloads surrounding AI initiatives are manageable, which is especially crucial for a team of about four team members.
“There are a lot of balls in the air and several workstreams running at once, so people are encouraged to work in discrete blocks of time and limit context switching,” Lead Data Scientist Jeremy Biggs said.
Below, Ling, Mitchko, Biggs and five other AI experts share how they set direction for responsible innovation on their team, and what job candidates should know about how their team balances strategic direction with experimentation.
Featured Companies
Capco is a global management and technology consultancy that serves organizations in the financial services and energy sectors.
What’s the leadership like on the AI and machine learning team at Capco?
We are hands-on leaders who combine business acumen and technical understanding, and integrate AI tool usage heavily into our day-to-day tasks and workflows — all to achieve the best outcomes for us and our clients.
How does leadership help your team decide where to focus AI or ML efforts?
All good ideas are welcome, and teams are encouraged to stay on top of AI news and community groups for emerging patterns. All potential use cases and ideas are tracked in a common idea funnel, which goes through prioritization with support from business domain leaders and tech/AI leaders before AI governance processes.
“All good ideas are welcome, and teams are encouraged to stay on top of AI news and community groups for emerging patterns.”
What should candidates know about how your team balances strategic direction with experimentation?
We experiment constantly and fearlessly because we believe the status quo is changing rapidly, and that good strategic direction will come from ideas that have been tested for desirability, feasibility and viability before doubling down on investment to scale.
SharkNinja develops a broad range of home products, from hair dryers and vacuum cleaners to blenders and air fryers.
What’s the leadership like on the AI and machine learning team at SharkNinja?
I’m specific about the outcome and flexible about the path. Direction is the part I hold tightly. Every AI effort has to name the business outcome it changes and who owns that outcome once the work ships. If we can’t answer that, we don’t start.
The path is where I step back. I give people real autonomy to lead in the places where they’re most capable and most comfortable because that’s where they move fastest, and I stay close enough to be useful when it counts. Then I keep pushing on the edge of that. The work I’m proudest of has come from somebody taking a problem slightly bigger than the one they raised their hand for.
The last piece is how close we sit to the business; we are not a lab. The people who will use what we build are in the room while we build it, which shortens the distance between an idea and an honest reaction to it. That proximity is also the guardrail. It’s harder to ship something careless when the person who has to live with it is reviewing it next to you.
“It’s harder to ship something careless when the person who has to live with it is reviewing it next to you.”
How does leadership help your team decide where to focus AI or ML efforts?
We start with the business decision and work backward from there. AI is always in the mix when we’re solving a business problem, and it should be. What we’re careful about is the order. We name the decision that actually has to be made, and then we ask honestly whether AI or machine learning is the right way to make it, rather than starting with the technology and going looking for a problem to attach it to. AI isn’t a shiny object we’re chasing.
A recent example is the dashboard we built for star ratings and social listening, where we deliberately left AI out for now. The processing logic and the engineering behind it had to be verified and reviewed by people, because the entire value of that dashboard is that the numbers can be trusted. We’ll layer AI in later as a feature once that foundation is solid.
That thinking shapes how we measure, too. We hold the work to the business outcome it was meant to change, rather than to model specs or confidence scores. Those tell us the model is performing well, not that the decision got better.
What should candidates know about how your team balances strategic direction with experimentation?
We hold the direction steady and leave the path to it genuinely open. Most of what we do begins as a pilot, and we’d rather run two or three different approaches to the same problem than commit early to whichever one sounded best in a meeting, because the version that works is rarely the one we predicted. We keep those pilots small and short on purpose, agreeing up front on what we’re testing, how long we’ll give it and what would tell us to stop, and we settle the data questions before anyone starts building.
What matters most is what happens when a pilot doesn’t work because we’re comfortable saying we were wrong and we try to say it early, while changing course is still cheap. Pivoting isn’t a failure here, but the reason we pilot in the first place, and what we try hard not to do, is let something limp along because it was already announced.
For anyone joining, that means you’ll have real room to try things, and an experiment that doesn’t land won’t be held against you as long as it was a real question and we learned the answer. What we do ask is that you can connect what you’re exploring back to a business outcome that somebody cares about.
InterSystems’ cloud-first platforms enable organizations from various industries, such as healthcare and financial services, to power their applications with clean, accessible data.
What’s the leadership like on the AI and machine learning team at InterSystems?
AI leadership at InterSystems is collaborative, practical and enablement-first. Our job is not simply to approve tools or set policies. It is to help teams turn good ideas into measurable outcomes by giving them access to the right technology, expertise, reusable components and governance. We work closely with technical teams, business leaders, domain experts, legal, security and data protection because successful AI requires more than a strong model. It requires trusted data, thoughtful implementation and clear human accountability. We encourage people to use AI boldly, while remaining responsible for the quality and impact of their work. The goal is to create an environment where teams can move quickly without compromising the standards our customers expect from InterSystems.
“We encourage people to use AI boldly, while remaining responsible for the quality and impact of their work.”
How does leadership help your team decide where to focus AI or ML efforts?
We begin with the problem and the desired outcome, not the technology. Leadership helps us prioritize opportunities based on business value, user need, technical and data readiness, risk and the potential to reuse a solution across teams. We look for initiatives that can improve productivity, quality, customer outcomes or employee experience, then establish success measures before scaling them. We also connect domain experts with technical builders early, so solutions reflect how work is actually performed. Small, measurable experiments help us test assumptions and collect evidence. Initiatives that demonstrate value can receive additional investment, while lessons from less successful experiments are documented and applied elsewhere. This keeps the portfolio focused without discouraging creativity.
What should candidates know about how your team balances strategic direction with experimentation?
Candidates should expect both clear direction and meaningful freedom to explore. We have strategic priorities, shared governance and defined standards, but we do not believe innovation can be planned entirely from the top down. Many of the best opportunities are identified by employees who understand a workflow or customer problem deeply. We create safe spaces for those ideas through pilots, hackathons, evaluations and close collaboration between domain and technical experts. The expectation is that experimentation produces learning, not just demos. Teams should test with realistic use cases, evaluate results critically, share reusable work and maintain human ownership of the outcome. Curiosity is highly valued here, particularly when it is paired with rigor, responsibility and a willingness to turn an experiment into something others can trust and use.
360Learning’s AI-driven platform is designed to enable learning and development teams to create content, automate tasks, boost engagement and more.
What’s the leadership like on the AI and machine learning team at 360Learning?
At 360Learning, leadership is about creating the conditions for people to own decisions. We operate with low authority and high accountability: Teams decide how to achieve an outcome and remain accountable for its impact.
“We operate with low authority and high accountability: Teams decide how to achieve an outcome and remain accountable for its impact.”
This is especially important in AI engineering, where it’s impossible to know upfront which ideas will work. Leadership isn’t about having all the answers; it’s about providing strategic context, defining measurable outcomes through objectives and key results, and giving teams space to test assumptions and learn quickly through a focused proof of concept, a small client test or a low-cost experiment.
We actively encourage innovation: Teams are empowered to challenge the status quo, explore new approaches, experiment and learn from failure, while staying focused on customer value and business impact. We also challenge decisions with facts. Teams bring data, explain their reasoning, track impact and consider alternative perspectives, especially when results don’t meet expectations.
Ultimately, leaders set the context, priorities, guardrails and ambition; the people closest to the problem own the decision and have the freedom to innovate their way toward the outcome.
How does leadership help your team decide where to focus AI or ML efforts?
Leadership helps the team prioritize efforts through our company objectives and roadmap, while balancing customer impact, business impact and innovation.
Our high-level roadmap is aligned with our strategic objectives for the year. This gives us a clear direction and shared understanding of where to invest our effort, without rethinking priorities every quarter. When evaluating an AI or ML opportunity, we look at the value for our clients, the impact on the business and the effort required. We want to focus our resources where the expected impact is highest, whether that means improving the product, solving a customer problem or creating business value.
At the same time, AI evolves too quickly for everything to be planned upfront. We leave room to test new technologies and ideas through small experiments, without immediately committing significant resources. Then, we can see if it can lead to a further investment and a place in the roadmap.
What should candidates know about how your team balances strategic direction with experimentation?
Candidates should know that we expect both strategic thinking and a strong appetite for experimentation. The strategic direction gives teams a clear understanding of what we’re trying to achieve, but it doesn’t prescribe every solution. We want engineers to challenge assumptions, explore new approaches and test what new AI capabilities can actually bring to our product/business.
The key is how we experiment. We favor small, focused bets that allow us to get a signal quickly rather than investing heavily before we know whether an idea works. A POC, a limited feature or a test with a small group of clients can be enough to learn whether something is worth pursuing.
When the signal is weak, we should be comfortable stopping or changing direction. When it’s strong, we give the team room to move quickly and invest further.
For candidates, this means a high degree of ownership: You’re not expected to wait for permission to test an idea, but you are expected to connect your experiments to a meaningful problem, measure the impact and learn from the outcome.
CodePath.org partners with universities and tech organizations to offer computer science students industry-vetted courses and career support so they can become AI-native technologists.
What’s the leadership like on the AI and machine learning team at CodePath.org?
CodePath's AI and machine learning team is relatively small, consisting of four team members. The director leading the team still writes code and contributes to analyses.
Leadership on the team is built around responsibility and ownership. The team sets goals each quarter, and every member of the team owns and leads a set of them, with the trust to plan that work and deliver it without asking permission along the way. There’s a weekly meeting to check status and clear roadblocks, and this the main avenue for director oversight, along with weekly one-on-ones with team members.
The operating belief is that the job is to hire A-players and let them do their best work, so direction gets set at the level of goals rather than tasks. Experimentation is encouraged, as long as it’s pointed at one of those goals. CodePath is a nonprofit with a startup culture, and pivoting quickly when something isn’t working matters for a team this size.
“The operating belief is that the job is to hire A-players and let them do their best work, so direction gets set at the level of goals rather than tasks.”
Where leadership does step in is on workload. There are a lot of balls in the air and several workstreams running at once, so people are encouraged to work in discrete blocks of time and limit context switching.
How does leadership help your team decide where to focus AI or ML efforts?
Each quarter, the team director works with other members to set goals that they will be accountable for. They cover two kinds of work: serving the rest of the organization with internal reports, dashboards and the data infrastructure underneath them, and building CodePath’s evidence base through research partnerships with outside evaluators.
Where machine learning fits gets decided in that same conversation, rather than on a separate track. Once a goal is set, the next question is how the deliverable actually gets built, and whether a model is the right way to build it.
The clearest case is classification. Large language models sort student outcomes data at a scale nobody could review by hand, and they apply the same rules on every run, which manual classification never managed.
We generally set two hard limits on where AI can be used. It can be a second set of eyes reviewing code changes but can never be the primary “reviewer” of any code, especially code that was generated by AI agents. And any analysis that could be visible inside or outside the organization doesn’t get signed off without a person first reviewing it.
What should candidates know about how your team balances strategic direction with experimentation?
The quarterly goals provide the strategic direction for the team, but staff have a lot of latitude in how they get there. We don’t decide in advance exactly which tools or methods will be used to achieve a goal. If there’s a better way to do something, people are expected to try it.
The important part is knowing when to stop. We don’t want experimentation to become work in its own right, so it needs to be connected to a specific question or deliverable. If an approach isn’t working, the expectation is to move on rather than keep investing in it because we’ve already put time into it.
That can mean changing course several times during a quarter. The team is small, the work is varied, and there are usually several reasonable ways to solve a problem. Staff are expected to have substantial freedom over how they approach their work, but also to make decisions about what is worth pursuing and what isn’t.
Gradient AI serves clients that specialize in group health, property and casualty insurance and workers’ compensation, offering AI solutions that perform tasks like predicting underwriting and claim risks and reducing quote turnaround times and claim expenses.
What’s the leadership like on the AI and machine learning team at Gradient AI?
Leadership is split between developing AI/ML as part of our core product offering, which I lead, and the more general use of AI for everyday productivity across all our employees, which is a function that falls under our CTO. As AI becomes increasingly commoditized, these lines have started to blur. For example, I have led R&D to summarize customer call transcripts in the past, and now that is a standard offering in nearly every application managing text data. Good leadership must work together to define what is a core competency of our company requiring deep investment in R&D, and what is a commoditized or generally available feature which is better sourced externally, whether to enable our core products or our employees. Doing that well at a leadership level requires deep expertise in AI science and technology to be able to build what you need, broad experience across markets and product leadership to know how to buy what you can, and the wisdom to know the difference.
How does leadership help your team decide where to focus AI or ML efforts?
The best leaders explicitly do not decide where to focus AI or ML efforts. The best leaders decide where to focus overall investment to solve the problems that need to be solved. At the end of the day, our job is to create a product that creates value for our customers, and the most important part of that job is to focus on the value proposition and the customer need. This is what we call “customer-obsessed.” Exactly how we solve these problems — with AI, with traditional data science/machine learning techniques, with human-in-the-loop or even rote automation — is an implementation detail.
“The best leaders decide where to focus overall investment to solve the problems that need to be solved.”
My role as an AI/ML leader is to help envision the “art of the possible” with leadership and create possibilities for our customers, but an equally important part of my role is to reframe the problems as “non-ML” for leadership so they can solve them in more appropriate ways, with more appropriate teams. Leadership decides what to build, not how to build it, and my role in collaboration with other technical leads is to enable that with the best possible outcome and ROI.
What should candidates know about how your team balances strategic direction with experimentation?
I like to think about “strategic experimentation.” Experimentation serves many purposes: It sometimes unlocks revolutionary tech, but sometimes, it’s needed for team bonding or learning opportunities. Many teams “experiment” without being strategic about what they want to accomplish. This can lead to disappointment when there is a mismatch between what the employees and the company want to get from the time. Like everything, better results come when we act strategically and with purpose.
To help my team grow as leaders, I set a blanket five to 10 percent time allowance for any experimentation that is clearly justified by our goals. Want to try a new technology, read a few long-shot papers or try a new code design? It’s your time to invest, and your leadership is demonstrated by how well you manage these bets and if they consistently pay off or not. Other techniques, like hackathons, can also be successful if there is strategic alignment; a hackathon intended for team bonding will be very different from one intended to try five moonshot techniques and decide which are worth pursuing. Being clear about what we are trying to accomplish makes a huge difference in the impact we drive.
Jasper’s AI-powered workspace helps marketers handle a variety of tasks, such as product marketing, PR and performance marketing.
What’s the leadership like on the AI and machine learning team at Jasper?
We don’t have a dedicated AI/ML team, and that’s intentional. AI work lives inside every team here. Each team owns both the AI products and features we ship to users and the ongoing shift in how we build software. So, the engineer fixing a bug today is also thinking about whether that class of bug should be handled automatically next time. We focus our time on changes that materially improve the product or the way we work, such as automating parts of feature development and bug resolution, or catching and fixing issues before they become bigger problems. Leadership’s role is mostly to set direction, create focus and make decisions about where we should place real bets. A few of us, along with our developer experience team, spend more dedicated time on this, but the ideas come from everywhere. That distributed ownership gives us a much wider set of possibilities to work from, while clear prioritization lets us quickly set aside the ones that aren’t worth pursuing and put real weight behind the ones that are.
“Leadership’s role is mostly to set direction, create focus and make decisions about where we should place real bets.”
How does leadership help your team decide where to focus AI or ML efforts?
A lot of leadership’s time goes into figuring out where we need to be in a few weeks, a few months and further out. We hold those views loosely enough to adapt as AI capabilities change, but firmly enough that teams know what they’re building toward. The practical effect is that we invest heavily in strong core systems. If the foundations are solid, a new model, architecture or capability isn’t a disruption; it’s something we can plug in, test and learn from quickly. The AI space can look like a roller coaster from the outside. Internally, it looks more like a queue of new things we get to experiment with because the underlying systems are built to absorb change. So, when a team is deciding where to focus, it usually comes back to two questions: Does this move us toward where we believe we’re headed, and does it build on or improve the foundation we’re going to need to get there? If the answer to both is yes, we move. If not, we step back and discuss before committing time to it.
What should candidates know about how your team balances strategic direction with experimentation?
Strategic direction comes first. We want teams to understand what we’re trying to accomplish for users and make decisions in that context. Direction can change, especially in a space moving this quickly, but when it does we shift together rather than having individual teams drift in different directions. Experimentation is how we get there, not a side activity. Experiments pointed toward where we’re already going are how we learn faster, uncover better ways to solve problems, and sometimes get ahead of needs customers haven’t fully articulated yet. So, a candidate should expect real room to try things, and should also expect to be asked how those experiments connect to where we’re headed. If you like both halves of that, you’ll probably do well here.
Energy CX is an energy brokerage that aims to help its clients maximize savings, efficiency and sustainability.
What’s the leadership like on the AI and machine learning team at Energy CX?
Our approach is simple: empower, experiment, measure.
Empowerment comes first because the best AI use cases don’t originate with me — they originate with whoever feels the friction. One salesperson built himself a sales dashboard to streamline his workflow, and another built an automation that sends quotes directly to customers. Neither was assigned; both were people closest to a problem who were given the tools and the permission to solve it. To help encourage innovating with AI, we ran an internal artifact challenge. We wanted to make that behavior visible and reward it.
“Empowerment comes first because the best AI use cases don’t originate with me — they originate with whoever feels the friction.”
Experimentation is the default posture rather than a stage we schedule. I believe it’s better to ship 10 small things and learn from the three than spend a quarter specifying one.
Then, we measure. Enthusiasm isn’t evidence. If a workflow didn’t save time, improve accuracy or unlock something we couldn’t do before, we say so and kill it. That discipline is what keeps the empowerment from turning into a pile of half-used prototypes — and it’s what gives the team room to keep experimenting.
How does leadership help your team decide where to focus AI or ML efforts?
Mostly by lowering the cost of finding out. I run weekly AI office hours. It’s part news about what changed in the AI world in the last seven days and what’s becoming possible, and part clinic, where anyone can bring a piece of their day-to-day work, and we work on it live. Recent sessions have covered what skills are and how to build them, automating recurring work with scheduled tasks and routines, and using artifacts to manage work in the databases we use.
The effect is that prioritization happens bottom-up with guidance rather than top-down with a mandate. I’m not guessing at which workflows are painful; people tell me, weekly, with specifics. Where I do set direction explicitly is on unstructured data and intelligence. In energy procurement, most of the signal lives in contracts, bills, call transcripts and supplier correspondence, text that was economically unreadable at scale until recently. That’s the highest-leverage frontier for us, so leadership has prioritized deliberate investment rather than opportunistic attention.
What should candidates know about how your team balances strategic direction with experimentation?
First, you’ll have real latitude. We don’t route AI ideas through a committee. If you see something worth trying, you’re expected to try it, not to write a proposal about trying it.
Second, you’ll be asked to measure it, and the bar is specific. Before you build, define what improvement looks like. Then, answer the question I ask about nearly everything: Is the declining cost of intelligence applicable here? If a task was too expensive to do well a year ago and is now cheap, that’s where we should be spending our attention. And if the answer is no, this doesn’t get better as models get cheaper and stronger, I want to hear why not. That’s usually the more interesting conversation because it tells you the constraint is data, process or trust rather than capability.
Candidates who thrive here are comfortable being given a direction instead of a spec, and comfortable having their own work evaluated honestly. If you want your ideas taken seriously, this is a good place to bring them.
