Building the Labelbox Team

Nov 01, 2019

At Labelbox, we're not just product builders, we are team builders too. Behind the scenes, we invest a significant portion of our time attracting and recruiting talented people. It's a top priority. And we've made progress: in the past 4 months, we've grown our team 75%! We're proud of everything Labelbox has accomplished to date, but we realize we need more great people to join us in order to fulfill our vision:

Build the best products for humans to advance artificial intelligence.

While building a great team is an ongoing investment, we want to take a minute to introduce a few new faces and the capabilities they bring to the team.  

Where we are investing

Backend Engineering: Building world-class AI infrastructure

Kyle Nesbit

Kyle Nesbit is the first Staff Engineer at Labelbox. Kyle's expertise is in backend infrastructure, distributed systems, databases and machine learning pipelines. He holds 9 patents in the area of high-performance distributed systems, including the tech that powers Google Big Query and Google Cloud Dataflow. He brings a wealth of knowledge and experience as we design Labelbox for world-class reliability and scale.

Some of our most advanced customers utilize realtime AI training data workflows. This means that sensor data is captured, labeled by the model, and reviewed by a human, all within a few minutes. Downtime and performance issues could cost our customers thousands of dollars in lost labeler productivity and downtime for their products and services.

In order to serve these use cases, we started to think of Labelbox as critical infrastructure. Kyle helps our team as we build out capacity for the future to handle more of these customers who not only need realtime workflows, but massive scale and global responsiveness.

Frontend Engineering: Creating the label editor of choice

We wish we had a picture of Alex, but he's become a generative art piece at this point.

Alex Bainter is part of our globally distributed engineering team. While we could talk endlessly about Alex's qualifications, all anyone seems to chat about is his generative music. If you're not familiar with the concept, two criteria must exist for music to qualify as “generative":

  1. It must change continuously and never repeat itself exactly.
  2. It must last forever.

Alex and a cadre of other developers joined Labelbox to expand the functionality of our label editor (our nomenclature for the software where one labels training data). We're constantly seeking ways to improve labeling efficiency both algorithmically and through improved ergonomics for labelers; we also invest heavily in expanding functionality for the broad applications of computer vision.

There's a lot of interesting work we can't talk about yet, but we can't wait to show you what Alex and the rest of the engineering team are working on.

Docs: Ease of use for developers

Alex Cota

Alex Cota is one of those rare people who we knew immediately was right for the job. She brings a wealth of experience as a technical writer for perception systems at Cruise Automation where internal engineers relied heavily on her documentation.

We place strategic priority on being the "choice of developers." Great, accessible, and intuitive docs are the backbone of any blissful developer experience. In the long run, we believe that machine learning development is an iterative process and that as data science teams mature, they need software to manage their training data. Alex is leading the company in the creation of a world-class docs experience for the Labelbox community.

Product Design: Bringing craftsmanship to every product we make

David Kjelkerud

David Kjelkerud was once mistaken for a Euro indie-film director. He brings thoughtful insight to our product decisions beyond design, extending his talents into product management and strategy as well. David led a team of designers at Dropbox where he earned his street cred on Dropbox Paper.

There is a philosophy of craftsmanship in everything we do at Labelbox, in fact, it's one of our core values:

Craftsmanship is the fine balance between perfection and a solution to the given constraints. To be a craftsman is to be consumed in a problem and to create your best expressions of work as a result. You can't be a craftsman in every thing, you can only be a craftsman in a few things, but you do those things remarkably.

David lives at the intersection of craftsmanship and design, ensuring that everything we do is both functional and beautiful. To illustrate his focus, David will be flying to Asia next month to sit with our labelers and work side-by-side, labeling, for a full 8-hour day, just to see how our users experience the software. If that's not commitment to craftsmanship, we don't know what is.

Labelbox Workforce: A complementary, high-quality labeling service

Audrey Smith

Audrey Smith joins us from Amazon A9 where she was responsible for machine learning labeling operations: overseeing a portfolio of labeling service vendors and providing program management for Amazon's computer vision projects.

This year we introduced a new complementary service, Labelbox Workforce, which allows our customers to purchase add-on, data-labeling services in addition to our software platform. Customers love working with Labelbox and asked us for a high-quality alternative to dealing with some frustrating, low-quality, and expensive third-party vendors.

With the ability to aggregate demand from a number of customers, Labelbox is able to secure partnerships with the highest quality data-labeling services. When paired with the quality and efficiency gains produced by our software, the combination has been a "no-brainer" for our customer base. We're excited to have Audrey's expertise leading the expansion of this offering.

Marketing: Creating a new category for AI training data software  

Craig Hosang

Craig Hosang is leading our efforts to build out marketing. He joins us from GitHub where he helped define the company narrative and led their enterprise product marketing efforts prior to the acquisition by Microsoft.

We're still in the early days of standardizing the AI training data workflow. We asked Craig to help us create our category and drive consensus in the market around the need for high-quality training data. It's a tall order! Yet, overlooking the need for investment in high-quality training data is one of the major stumbling blocks companies hit trying to get their AI ambitions into production.

Labelbox has been incredibly blessed with strong organic growth to date; we're looking to Craig to help us move forward and share our story with a broader community of people.

Some fun elements of our emergent culture

College friends make for deep connections

There are a number of friends at Labelbox, especially amongst the early team members.

Manu (second from the left), Brian (second from the right), Cyrus (farthest to the right)

Manu Sharma, Brian Rieger, and Cyrus Jou all attended Embry-Riddle Aeronautical University where they studied aircraft design. In fact, the three met in the same class, competing to produce the best new airplane design. Thoughtfulness from their days designing airplanesengineering where lives mattertransfers into the attitude these three bring to the office.

Alex Pattison and Connor Harwood, and Kyle Owens and Ryan Quinn are "package-deal" engineers. Both pairs went to college together and bring an instant community to the engineering team.

You can listen to Manu (CEO) and Brian (COO) talk about how they met, Labelbox's origin story, and where we're going in a recent podcast interview with Eye on AI.

Culture of electronic music

Whether intentional or not, Labelbox has a strong culture of audiophiles and electronic music lovers. We joke every time a new hire mentions this interest as it's beginning to look like a hiring requirement (it isn't, we swear!). While we don't get to spin often, we have a DJ booth in the office and there are a number of talented DJs and DJs in training amongst our team.

You can check out one of the DJ sets on the official Labelbox SoundCloud. Fun fact, Labelbox currently occupies SoundCloud's former offices in the Mission District, San Francisco.

The journey forward, together

We're super excited and proud of the team we've built; everyone here's committed to serving our customers and humankind overall.

We're on a mission to build the best products for humans to advance artificial intelligence.

You can expect more great things from us on the horizon. Stay tuned.

Labelbox

Labelbox is a collaborative training data platform empowering teams to rapidly build artificial intelligence applications.