Community Spotlight: Baidu Ventures

Prior to Baidu Ventures, Fang Yuan worked at Comet Labs, an early stage AI & robotics fund, as an Operations Director. Prior to Comet, she worked in business strategy & operations for two “unicorn” startups. Early in her career, Fang was a consultant at Oliver Wyman Financial Services consulting, focusing on general management, operational and IT improvements for major corporations globally (U.S., Canada, Brazil, Mexico, Nigeria, etc.).

Let’s start with some background on Baidu Ventures. Can you tell us a bit about your fund’s mission and its AI-focused approach?

Baidu Ventures is the non-strategic investment arm of Baidu, launched in early 2017 with a $200MM first fund. Our particular focus is on early-stage (Seed & Series A) AI & robotics, applied to specific industry verticals such as healthcare, transportation, agriculture, retail, etc. We’ve made ~45 investments outside of Asia, which is the region my team covers from San Francisco.

Our mission is to work with the best entrepreneurs to transform and reconstruct the world through AI & robotics. Personally, I’ve immensely enjoyed getting to know all of our portfolio founders and seeing them grow their companies over the last few years — it’s been quite inspiring.

Recently, Baidu Ventures hosted an event in collaboration with Silicon Foundry and SK hynix that targeted AI product managers. Why do you think this is an interesting community to engage with?

My co-host Melody and I really enjoyed hosting our first AI Product Manager event and look forward to doing more with this community. There are a number of reasons for why it’s an interesting community. The first being that before we hosted this event, there really was no overarching community for people managing AI products; there were scattered, smaller groups and listservs of people doing this type of work, but they weren’t being brought together. I think there’s always value in bringing together people working in a new field to learn and share stories, promote emerging best practices and more.

The second is that there’s a real reason for AI product managers to come together, and we saw that in the overwhelming number of responses to our call for applications to our event. AI product management is a nascent field and everyone is constantly finding better ways to gather and prepare datasets and use them to train models. There’s also a lot of trickiness around deploying and troubleshooting models, and then monitoring them afterwards. Not to mention, understanding how to set up data sharing agreements with data sources if that’s relevant to their business.

There’s no standard set of best practices yet around creating, maintaining and innovating on AI products, and we’re excited to bring people together who are advancing this field. We are able to help each other with insights as to what’s working or not within their own companies.

What are some of the challenges and opportunities you heard from that evening’s conversations?

We had a panel of four AI PMs from both large and small companies and all with very different products. The diversity in our group was great in that it gave richness to the variance of thought, experiences and responses.

Since we agreed to adopt the Chatham House Rule for the session, I can’t give specific details of the conversation. However, on a macro level, the big challenges discussed were around data privacy and ethics, and how to manage that with building the best product as quickly as possible, working with the set of sometimes fragmented tools required for managing, deploying, and monitoring models as well as managing user expectations.

The opportunities are massive in that we’re still at the very early stages of what will be possible with AI & robotics. There was a lot of optimism in the room around what might be possible in the coming years, especially as the tooling and foundational infrastructure for AI & robotics become more mature.

More broadly speaking, how do you see the AI landscape changing in 2020?

I think from an investor’s perspective, it’s not so much how the AI landscape is changing, but rather how the adoption of AI technologies is or is not happening. It’s not so much the technology, but rather how it’s applied that counts.

For example, I spend a lot of time with corporates to understand their business challenges and needs because all of our startup investments are enterprise-facing, so it’s vital to understand the customers’ perspectives. It’s very clear that corporates are starting to deeply think through their AI strategies, with some being more advanced than others, but the majority are still at early stages.

As part of that evaluative process, corporates are much more open to working with startups than before, which I think will allow a generation of startups to mature, as they sell into that ecosystem. No large company wants to get left behind as competitors start adopting better technologies. I think we will start seeing AI-first startups that will grow quite large as they develop products to address the needs of Fortune 500 companies. The key is to find a niche, but high-value, problem to start off with and then to expand from there.

As venture dollars being poured into AI startups continue to grow, do you see Baidu Ventures’ approach toward AI investing changing?

I think our strategy isn’t so much changing as it is deepening. We’re committed to only focusing on AI & robotics startups in the enterprise space, and each of us on the team spends a good portion of our time deep diving into specific technologies and industry verticals so that we can more easily know what we’re looking for, be able to ask the right questions and also be a good partner to the founders who we invest in. In other words, we believe specialization is a critical differentiator that will make us better investors.



Get the Medium app

A button that says 'Download on the App Store', and if clicked it will lead you to the iOS App store
A button that says 'Get it on, Google Play', and if clicked it will lead you to the Google Play store