_Our new Executive Function series features perspectives from leaders driving transformation through AI._
DoorDash is one of the world's leading local commerce platforms that helps businesses of all kinds grow and innovate, connects consumers to the best of their neighborhoods, and gives people fast, flexible ways to earn. We spoke with Mariana Garavaglia, Chief People Officer about measuring AI literacy, using AI to augment human judgment, and how it is helping employees make an even deeper impact.
#### You’ve previously described HR operating more like a product team or R&D lab. How has AI enabled or accelerated that kind of transformation at a company like DoorDash?
First, it helps to understand how we operate at DoorDash overall—it’s an operator culture.
Everyone is expected to deeply understand the consumer experience and move with the speed and mindset of a product team. Long before AI advances, we were already constantly running experiments—from conversion rates to marketing campaign efficacy. AI now supercharges that by accelerating how quickly we can learn, test, and iterate.
Tooling is a core responsibility for my HR and IT teams, and AI plays a key role across three layers. First, we focus on access and literacy—ensuring everyone, including non-technical roles, can use frontier tools like co-pilots through enterprise rollouts, hackathons, and tutorials. Second, we integrate internal data to break down silos, enabling AI-powered search and smarter content delivery. Third, we’re exploring where AI agents can take on tasks in a smart, trusted way. The goal is not just automation but also enhancement of core workflows that improves the way our employees experience their work.
As a People team, we believe it’s important to raise both access and literacy around AI across the org. We’re responsible for helping employees be as productive, engaged, and effective as possible across their experience and workflows. It’s an incredibly exciting time—our work impacts both our internal team and the broader enterprise.
#### What AI tools or capabilities have been most impactful to your operations or for employees day-to-day?
Even enterprise chat tools have had a huge impact. What’s really exciting is how they’ve enabled more employees to become technical creators. For example, someone on our people ops team built a script to automate document uploads—something they used to do manually one by one. That’s a huge shift. It democratizes automation and workflow creation that used to require engineering help.
These changes are hard to measure in the moment, but even small examples show how AI can empower individuals to drive value in their workflows. Now, non-engineers can quickly build solutions tailored to their use cases.
> “Across all teams, we're super excited about how AI is enabling us to accelerate building these experiences. It really accelerates the speed with which we can learn and test and iterate.”
How do you measure AI literacy or impact as you scale adoption?
We’re still early, but we’ve started with two foundational metrics: adoption and frequency of use. We know engineers can now build in minutes what used to take days. And adoption has been fast and organic. Teams are naturally solving problems using the tools available to them.
We’re also thinking about distribution—who has access to licenses, how evenly they’re distributed, how frequently they’re used. Beyond that, we’re starting to integrate AI literacy into our performance framework—thinking about competencies like willingness to adopt new tools and a learning mindset.
Each function is measuring impact at its own pace. Engineering teams are ahead. Other functions will get there over time. The tools are evolving quickly, so we’re starting with durable metrics and expanding from there.
#### How should HR leaders think about how AI is changing employee experience and engagement?
While a lot of the focus has been on delivering services more efficiently, I’m really excited about how AI can enable people to grow and drive more impact in their work.
We’re seeing how non-technical employees can now create personalized content, automate workflows, analyze data—things that used to require technical skills. That’s huge from a learning and development standpoint. DoorDash already has a strong culture of self-directed learning. AI lowers the barrier even more, enabling more mobility and tailored development.
Eventually, we’ll be able to offer personalized, scalable employee development plans, customized to each person’s goals, performance patterns, and career trajectory. It’s a shift from one-size-fits-all to truly individualized support.
> “We now have a tool that opens the door to much more personalized, much more scalable employee experiences, [and] development plans that are tailored to each individual's goals, to each individual's performance patterns, and to different role trajectories.”
#### You’ve emphasized that AI should augment—not replace—human judgment. Can you share examples?
Absolutely. Two great examples are performance reviews and employee surveys.
With performance reviews, our system collects a lot of feedback—but it's been hard to synthesize. AI helps us surface key themes, areas of strength, and areas for growth, so employees get clearer takeaways, not just raw feedback.
With surveys, we used to read thousands of responses manually. AI now helps us identify patterns and generate actionable summaries for managers. We've even built workflows that generate Mad Libs-style action plans tailored to individual managers, showing how their team responses have changed over time. It's boosted our people analytics work and made feedback more actionable.
#### Are you doing anything with AI around skill leveling, executive coaching, or upskilling?
We’re not using AI for executive coaching directly yet, but we are building predictive models to assess executive performance—both for internal promotions and external hires. We use cohort data, interview assessments, and reference checks to predict success. Executive coaching is one data point among many.
The idea is that AI can help us predict success with more confidence, and then coaching can be used to fill in any targeted gaps. We’re very clear that human interpretation is still essential—we use data and signals to support, not replace, human judgment.
> “AI now gives us the ability to move from reacting, learning—adult learning and development and one-size-fits-all programs—to much more individual growth paths, which I think is amazing.”
#### Looking ahead, what AI trends are you most focused on over the next 12–24 months?
For the HR team specifically, we’re still early in our agents journey, especially around core people workflows. That includes helping employees get policy answers, supporting development, enabling managers, and more.
The most exciting frontier is around personalization tech—how we deliver tailored experiences and support to every employee. That’s where we see the biggest potential.
#### Who on your team is responsible for building up the stack as you go deeper—from base tools to agentic capabilities?
We’re lucky at DoorDash that our HR and IT teams include dedicated engineers and engineering resources focused on internal workflows. Not every HR team has that, but it’s been a huge enabler for us. It lets us move faster and iterate more. It also means we can take a unified strategy and actually deliver it.
As you said, it’s key that we have technical support integrated into our people strategy. You need both the tools and the talent to move up the stack and fully realize the potential of agentic capabilities.
_DoorDash is using ChatGPT Enterprise across its organization in both technical and non-technical teams including finance, sales, operations, IT, and marketing. OpenAI APIs also power DoorDash’s customer service platform which serves 3 million chats per month, as well as internal workflows for review moderation and support._
Model ML is helping financial firms rebuild with AI from the ground up API Jul 23, 2025
AI powers Expedia’s marketing evolution B2B Story May 14, 2025
AI helps John Deere transform agriculture API May 6, 2025
Our Research * Research Index * Research Overview * Research Residency * OpenAI for Science * Economic Research
Latest Advancements * GPT-5.3 Instant * GPT-5.3-Codex * GPT-5 * Codex
Safety * Safety Approach * Security & Privacy * Trust & Transparency
ChatGPT * Explore ChatGPT(opens in a new window) * Business * Enterprise * Education * Pricing(opens in a new window) * Download(opens in a new window)
Sora * Sora Overview * Features * Pricing * Sora log in(opens in a new window)
API Platform * Platform Overview * Pricing * API log in(opens in a new window) * Documentation(opens in a new window) * Developer Forum(opens in a new window)
For Business * Business Overview * Solutions * Contact Sales
Company * About Us * Our Charter * Foundation(opens in a new window) * Careers * Brand
Support * Help Center(opens in a new window)
More * News * Stories * Livestreams * Podcast * RSS
Terms & Policies * Terms of Use * Privacy Policy * Other Policies
(opens in a new window)(opens in a new window)(opens in a new window)(opens in a new window)(opens in a new window)(opens in a new window)(opens in a new window)
OpenAI © 2015–2026 Manage Cookies
English United States
