On this interview, we sit down with Anirudh Reddy Pathe, Senior Director of Resolution Science at Glassdoor, to discover how data-driven decision-making is reshaping enterprise technique. From his journey spanning engineering, fintech, and journey to main determination science at Glassdoor, Anirudh shares sensible insights on constructing high-performing groups, aligning AI initiatives with enterprise outcomes, and fostering a tradition of experimentation. He additionally displays on classes discovered from failure, indicators of organizational readiness for AI, and the way the position of determination science could evolve with the appearance of generative AI.
Discover extra interviews right here: People within the Loop: 5 Voices Redefining the Way forward for AI
Your journey from engineering to main Resolution Science at Glassdoor is kind of dynamic. Are you able to stroll us by way of the pivotal selections or inflection factors that formed this trajectory?
My journey has actually been about following curiosity and leaning into moments the place I might have extra influence. Early in my profession, I used to be deeply technical, centered on engineering and problem-solving on the system degree. The inflection factors got here once I realized efficient utilization of knowledge can remedy vital enterprise issues at scale. At Priceline, I noticed the ability of speedy experimentation. At Uncover, I noticed how massive groups can transfer quickly collectively if a powerful tech staff and a powerful enterprise staff mix forces. At Glassdoor, I intentionally pivoted towards management in Resolution Science as a result of it sat on the intersection of technical depth and enterprise decision-making.
You’ve constructed and scaled international groups throughout corporations like Uncover and Priceline. What are your core rules for constructing high-performing information science groups in at the moment’s distant and hybrid environments?
I anchor on three rules: readability, autonomy, and connection. Readability round mission and success metrics so groups know the “why.” Autonomy so gifted individuals have room to unravel issues their means. And connection, particularly in hybrid setups, so nobody seems like a lone satellite tv for pc. I make investments so much in rituals that make distributed groups really feel cohesive, whether or not that’s shared OKRs, cross-team undertaking possession, or just creating area for casual interplay.
How do you make sure that AI and automation initiatives keep carefully linked to actual enterprise influence and never simply turn out to be ‘shiny objects’ inside organizations?
The query I all the time ask is: What consequence will this modification? If an AI mannequin doesn’t clearly tie to a lever the enterprise already cares about, income development, price effectivity, and person expertise, it dangers being shelfware. I push groups to outline success in enterprise phrases first, then scope the AI resolution. That self-discipline retains us from chasing shiny objects
With determination science touching so many capabilities, product, advertising and marketing, operations, how do you prioritize and steadiness cross-functional wants with out diluting focus?
Resolution Science sits in the midst of every thing: product, advertising and marketing, ops—and it’s simple to get stretched skinny. I’ve discovered to prioritize the place the flywheel impact is strongest. If an perception or experiment unlocks compounding worth throughout capabilities, that’s the place I’ll make investments. I additionally use clear roadmaps so stakeholders see trade-offs, which helps handle expectations
What do you see as the most important misconceptions leaders have about experimentation, and the way do you coach them by way of adopting a test-and-learn tradition?
One false impression I run into so much is the idea that each experiment has to ship a “win” to be useful. In actuality, most assessments don’t transfer the needle in the best way individuals anticipate, however that doesn’t make them failures. Each sharpens our understanding of buyer habits and guides us towards higher bets. One other false impression is that experiments are solely about UX tweaks or surface-level optimizations. The reality is, when performed proper, experimentation can affect strategic selections, what markets to enter, what merchandise to sundown, and how one can allocate funding. Serving to leaders see experimentation as a decision-making framework, not only a product instrument, adjustments the dialog completely.
Your work has spanned totally different industries from fintech to journey to tech. How does your method to data-driven development adapt throughout such assorted enterprise contexts?
Whether or not it was fintech, journey, or tech, the fixed has been utilizing information to scale back uncertainty and create worth. The difference is available in what threat and alternative appear like in every area. In fintech, threat was litigation and governance, however the alternative was immense to drive monetary outcomes for tens of millions. In journey, it was a conversion beneath intense competitors, and the chance was to supply memorable life experiences for tens of millions of individuals. At Glassdoor, the danger is person belief, however the alternative is the power to assist tens of millions of individuals discover the job they love. The playbook is similar: make clear the choice at stake, then deliver the correct mix of analytics, experimentation, and storytelling to affect it.
How do you see the position of Resolution Science evolving over the following 5 years, particularly within the wake of speedy advances in generative AI?
I feel Resolution Science will evolve into being the connective tissue between AI and the enterprise. With generative AI decreasing the price of evaluation, the differentiator shall be who can ask the correct units of questions and translate them into motion. 5 years from now, I anticipate Resolution Science leaders to be much less about constructing fashions/ evaluation themselves and extra about orchestrating people and machines to drive enterprise outcomes.
Are you able to describe a time when a data-driven perception conflicted with govt instinct, and the way you navigated that stress to drive alignment?
I’ve undoubtedly been in rooms the place the info contradicted intestine intuition. One instance: we examined a function executives beloved, however the experiment confirmed it really diminished engagement. Navigating that stress required empathy, acknowledging the instinct whereas being clear on the proof. In that case, exhibiting the longer-term influence by way of simulation helped construct alignment. The hot button is framing information as a choice companion, not a choice dictator
What indicators do you search for when assessing whether or not a company is really able to embrace AI-enabled decision-making at scale?
I search for three indicators: management alignment on the why of AI, a tradition that values testing over certainty, and information foundations which might be reliable. If any of these are lacking, scaling AI turns into a wrestle. You’ll be able to hack your means right into a proof-of-concept, however sustained influence wants these elements
For those who had been designing a graduate-level curriculum for Resolution Science leaders of the long run, what three programs can be obligatory and why?
Nice Query! I’d design it round:
Causal Inference & Experimentation – the spine of fine determination science.
Organizational Habits & Affect – as a result of insights don’t matter in case you can’t drive adoption.
AI Ethics & Governance – leaders want to grasp not simply what AI can do, however what it ought to do.
Inform us a couple of failure or false begin in your profession that in the end unlocked development or readability.
Early in my profession, I pushed for an enormous mannequin overhaul at Priceline with out totally aligning stakeholders. Technically, it was elegant; organizational, but it was useless on arrival. That failure taught me that affect and alignment matter as a lot as technical correctness. It modified how I lead ever since
- What’s a guide, framework, or piece of recommendation that essentially reshaped the way you lead groups or drive change?
- For those who needed to create a decision-making “dashboard” in your personal profession, what can be your high three metrics or indicators to observe?
If I constructed one for myself, the three metrics can be:
- Influence – am I shifting the needle for the enterprise and the individuals I lead?
- Studying velocity – am I stretching myself in new methods?
- Progress – am I capable of develop individuals on my staff
