Lenny Rachitsky | Lenny’s Podcast – Shweta Shrivastava (Waymo, Amazon, Cisco)
Shweta Shrivastava is a Senior Product Leader at Waymo, an autonomous driving technology company backed by Alphabet. Prior to joining Waymo, she was the CPO of Nauto, where she also worked on AI-assisted driver tools. Shweta has worked in product for over 15 years in senior roles at several companies, including Amazon and Cisco.
Trust is not a one-time thing. It’s a continuous thing.
Modifying autonomous vehicles for a more natural driving experience
The autonomous vehicle system is designed to adhere to speed limits for safety and efficiency, but it can lead to a less natural driving experience. The product team modifies the system behavior to provide a more natural driving experience, such as slowing down when driving downhill on a slope.
The team tracks commercial and operational metrics, such as the number of trips per week and daily active users, and driver performance metrics, such as safety and compliance with road rules. They aim to exceed human driving benchmarks for safety and avoid situations that require rescue help or slow down traffic for other users.
Prioritizing the customer and constant self-challenge
- When building a successful product or team, it’s important to focus on the customer or user problem and build something that creates value for them. It’s also crucial to prioritize and know what not to build, as trying to be all things to all people usually leads to failure.
- Large companies should avoid becoming complacent and should disrupt themselves before someone else does. This requires a mindset of constant self-challenge and transformational thinking.
- One effective practice for product managers is to write a press release for the finished product before even beginning to build it, which forces them to think deeply about the value proposition.
Using deep learning to create safe and natural driving behavior in self-driving cars
Waymo’s self-driving cars use deep learning models trained on human driving data to mimic human driving behavior in a good way, making it feel natural and safe for riders. The system can also understand signals like pedestrian body language and road signs, as well as social norms in different cities.
Compared to working in traditional software companies, working at Waymo is more technically complex due to the game-changing nature of fully autonomous driving systems, and project managers need to be comfortable with uncertainty and ambiguity.
The most important thing before product-market fit is to figure out what you’re not building. You have to be very ruthless about saying no.
The best way to build a successful product is to start with the customer problem and work backwards from there.