AdaptiLab

AdaptiLab was a Seattle-based software company that built tools to help employers evaluate and hire machine learning (ML) and data science (DS) talent. Its initial and primary product was an automated online technical screening platform positioned at the top of the hiring manager's interview funnel: it generated coding assessments customized to a company's domain and use cases, tested candidates on core ML competencies (data preprocessing, data analysis, feature engineering, model development/inference and general algorithms), added multiple-choice and short-answer questions covering theoretical knowledge, and automatically graded submissions to produce technical score reports and stack rankings of applicants.

The company described the assessment coverage as extending beyond machine learning engineers to most data-oriented technical roles, including data analyst, data scientist, business intelligence engineer, deep learning engineer and applied scientist. Beyond screening, the founders described plans for a technical diagnostic marketplace spanning sourcing, evaluation and placement — including access to pre-vetted candidates who had completed AdaptiLab coding challenges and learning modules — plus customized onboarding and training modules for new and experienced engineers. In July 2021 AdaptiLab sold the technical interviewing platform to CoderPad, which added it to its own interviewing product suite; the founders said they were building a new, undisclosed company targeting individual developers and development teams at companies and open source projects, backed by AdaptiLab's prior investors and new backers.

Founding story

Co-founders James Wu and Allen Lu were childhood friends who grew up in Seattle. Wu studied computer science and statistics at Duke University, worked on machine learning at two startups over roughly three years, was a 2017 Freestyle Capital engineering fellow and interned as a software engineer at AdStage. Lu studied computer science, machine learning and language technologies at Carnegie Mellon University, worked on deep learning teams at Google and Microsoft, and is a published deep learning researcher. Before AdaptiLab the pair built what they described as the most-viewed machine learning course on the developer-education platform Educative. They conducted more than 100 interviews with hiring managers and lead engineers, concluding that companies of all sizes and verticals made mistakes hiring ML talent, and used that customer discovery to define an MVP consisting of an interview environment with automatic code grading and a hiring manager dashboard for sending interviews and receiving score reports.

Business model

AdaptiLab sold software to employers for screening technical candidates, offering a hiring manager dashboard for dispatching automatically generated assessments and receiving automated score reports. The founders described an intended expansion into a technical diagnostic marketplace layered on the core screening product, covering sourcing, evaluation and placement of pre-vetted candidates as well as onboarding and training modules.

Traction

The first product version shipped at the end of March 2019, followed by regular feature releases. By September 2019 the team numbered four with plans to reach ten by 2020, and the product was in use by hiring managers at fast-growing tech companies and enterprises. Reported customers included Pinterest, Convoy and Experian. Total capital raised was reported at $2.2 million before the 2021 sale of the platform to CoderPad.

Latest developments

In July 2021 AdaptiLab sold its technical interviewing platform to CoderPad, an eight-year-old company that had raised funding from Summit Capital the prior year and that added AdaptiLab to its product suite; deal terms were not disclosed. Co-founder James Wu said he and Allen Lu were working on a new, undisclosed company targeting the developer market, with products for individual developers and development teams in companies and open source projects, supported by AdaptiLab's earlier investors plus new backers, and that a public launch was planned for later that year.