AI Studios with Natalia Burina Leading AI Teams | Sonny Patel Product & Engineering Exec (LivePerson, Amazon, Microsoft)
Sonny Patel is a tech executive who ran a team of 250 product managers and engineers at LivePerson. Before that, she ran a cross-functional org of designers, product managers, technical program managers and developers at Amazon. Sonny was also at Microsoft where she grew from an entry level product manager to a leader of leaders.
Here's a written summary of our conversation. You can also listen to the audio version via SubStack, Apple and Spotify. Or you can check out the convo on YouTube.
Rising to and Operating at Executive Level
Career Breakthroughs
I asked Sonny what was the pivotal moment that put her on the executive path. Sonny recounted how a VP at Amazon took a bet on her by entrusting her with a cross functional team. This expanded her scope significantly and set her up for bigger opportunities. Sonny attributes this breakthrough to both leadership support and a fortuitous situation. However, to position herself Sonny offered the following playbook.
Playbook
📖 Develop your knowledge and skills over years. No shortcuts here.
🏆 Champion - You need a champion to take a bet on you. Your knowledge and skills will enable your champion to stand behind you.
📍Situational Awareness - you need to recognize and capitalize on a unique opportunity to distinguish yourself. In Sonny’s case it was about saying yes to managing an engineering team in addition to product.
🎗Support System - You can’t make it without a professional support system.
Value of Good and Bad Managers in early career
Sonny believes that there is value in having one good and one bad manager early on. Watching and learning from other people’s management mistakes is a good way to build empathy for future reports. A good manager is able to provide psychological safety for their reports. This improves performance and sets people up to do their very best work. Sonny’s own experience with a bad manager taught her to have empathy and cultivate patience towards her reports. When people are new to product management or the AI space, managers should be especially patient.
How Sonny creates Psychological Safety for her Teams
🙋🏻♀️ Encourage and support people to ask silly questions
🙊 Allow people to make mistakes and learn without fear.
✔️ Sonny used regular check-ins and reporting mechanisms to monitor team progress and identify issues early
Building AI Products and Running AI Teams
How AI products are different
🪩 People tend to get enamored with the latest shiny technology. Sonny emphasized the importance of focusing on usefulness and not just the "cool" factor. AI Products should solve real problems for users in meaningful ways.
🔐 Privacy, Transparency and Control are critical. Users are willing to share data when they see a benefit and feel in control. Apply the idea of a privacy transaction when building products - if a product collects users data, the user should get something in return. Users should feel in control and everything should be done with their consent. Provide user control options in a coherent way that all fits together.
Why most AI products fail
AI products often fail due to edge cases that were not considered during design and testing. User expectations are often higher than what the technology can reliably deliver.
What makes Amazon an efficient execution machine
Before building a tech product, start with the customer and work backwards by understanding their problem. Amazon believes in the power of writing down things. Write a Press Release to imagine what your product unveiling may look like including all the related messaging. A Press Release is a one-pager that anybody should be able to read and understand. Some of the questions that a Press Release addresses are:
* What is the customer problem?
* Who is the target customer?
* Why is the idea big enough?
* Why now?
* What does the product development team say to customers?
* What would your customers say after using that product or feature?
* How is this overall fitting into your existing product strategy?
Furthermore a Press Release includes how the customers can get started, what they need to do, any associated costs, configuration experience, etc. After this, the team starts to dive deeper in terms of thinking about the product design aspect.
Sonny’s favorite aspects of the process are two things, Tenets and Rude Questions FAQ.
The Power of Tenets
Tenets are a set of principles around decision-making criteria. Having a clear set of tenets is useful for breaking debates during product design. Tenets define what is important in terms of trade offs. For example, sacrificing complex additional functionality in favor of simple and intuitive design for a non-tech audience. This is a potential debate that the product team could have. If the team was to make a trade-off, which side would they pick over the other? That's a great tenet. Definition of tenets requires a lot of thought.
Why you Need a Rude Questions FAQ for your Product
A rude Questions FAQ lays out all the difficult, unfair and rude questions that the team would rather not be asked, but might come up. Why? A Rude Questions FAQ will:
🔮 Prepare you for criticism. It's a crystal ball for future tough questions in key meetings and presentations.
🦾 It forces you to see your blind spots. We tend to become enamored with our own projects and cannot see our project's faults objectively. Rude Questions FAQ shatters any delusions.
📝 Help you build a Solid Plan - Collecting rude questions should take a while because you'll need to talk to a variety of people. It should also feed directly into your planning ensuring that you have a robust path to execution.
Organizational Mechanisms
There's a saying - "Good intentions never work, you need good mechanisms to make anything happen." A mechanism is a process to ensure that the team’s work is impactful with good outcomes on an ongoing, recurring basis. Amazon has set up organizational Mechanisms that make it incredibly effective.
What do Mechanisms do?
🧐 Allow the team to learn and improve from historical mistakes
⚙️ Create stable process-based solutions (rather than one off)
📑 Prevent tribal knowledge through Documentation
Sonny explained the mechanisms she set up for her team at LivePerson. Every other week she spent 90 minutes with all the engineering and product leaders spanning 15 teams. Each team had 5-7 minutes to present status. In each Meeting there were 3 situations that surfaced:
✅ Team was on track nothing needed
🟠 Team ran into solvable issues -> team would put together a plan to address the issue
⛔️ Off track, need help -> here Sonny could step in and get back on track
Where to find Sonny
Where to find Natalia
* YouTube
* SubStack
Generative AI for Business Workshop
🌟 I’m launching a Generative AI for Business Workshop in collaboration with Marily Nika, an AI Product Lead at Meta and formerly Google. We're organizing a three hour course held Friday June 30th - 9am-12pm PST and the cost is $99. The workshop will be recorded and you can access it at any point. If you're interested you can sign up here or click on the image below. Hope to see you there!
This is a public episode. If you would like to discuss this with other subscribers or get access to bonus episodes, visit aistudios.substack.com