AutumnTHE OLYMPIAD COLLECTION
Menu
RESEARCHED WITH AUTUMN
EB

Erik Bernhardsson

CEO at Modal
Modal Labs · New York, New York, United States · operator
Founder And CEO Of Modal Labs (serverless Cloud For data/ML, Founded 2021)CTO Of Better.com (2015-2021); Joined Among First 10 Employees, Grew Eng Team 2->300Engineering Manager At Spotify; Created Open-source Luigi And AnnoyHigh Frequency Trading Associate At Graham Capital Management (2011)IOI 2003 Gold Medal And IMO 2003 Participant, Representing Sweden

Erik Bernhardsson is a Swedish software engineer and founder based in New York City. He is the founder and CEO of Modal Labs (founded 2021), a serverless cloud computing platform for data and ML workloads. Previously he was CTO of Better.com (2015-2021), where he joined among the first ten employees and grew the engineering team from 2 to 300, and an engineering manager at Spotify (2009-2015), where he built the core music recommendation system and created the open-source Luigi and Annoy libraries. He holds an M.Sc. in Engineering Physics from KTH Royal Institute of Technology. Earlier roles include high-frequency trading at Graham Capital Management (2011) and internships at Google and D. E. Shaw & Co. He represented Sweden at the IOI (2002, 2003) and IMO (2003).

## Introduction I am currently working on various startup ideas. Until recently, I was the Chief Technology Officer at [Better](https://better.com/) where I managed a team of about 300 engineers. In the last 10+ years, I have focused on engineering management, recruiting, building consumer technology, machine learning, and math. My professional experience goes back to 1999, which was my first professional gig as a software engineer. ## Work experience Better Chief Technology Officer Feb 2015 – Jan 2021 New York City - Joined as one of the first 10 employees. - Took the engineering team from 2 people to 300. - Did an estimated 2,000+ interviews in the process of growing the team. - Migrated the legacy tech stack to AWS, Postgres, Python, and microservices. - Wrote code together with the team, building many critical components in Python and Node.js. - Built and introduced much of the core foundation for the data platform. - Introduced several machine learning models, including some with big impact on user acquisition and conversion. - Was in charge of a 5-layer engineering organization. - Managed and mentored engineers, promoted several people to managers and director

Details
LocationNew York, New York, United States
Company sitemodal.com
UniversityKTH Royal Institute of Technology
PythonC++C#Rmachine learningrecommender systemsdistributed systemsengineering managementmathsoftware engineering management
Notes
  • Quant employment verification: CONFIRMED. Erik Bernhardsson held a real, dated quantitative trading role: 'High Frequency Trading, Associate' at Graham Capital Management (Norwalk, CT), Mar 2011 - Sep 2011, per his own resume, which states he was one half of a team managing a $700M+ portfolio and implemented high-frequency trading strategies in C++ and C# for commodities and currency futures, researching strategies in Python and R. The role is distinct from competitions, camps, or offers. Employer classification: Graham Capital Management, L.P. is a real systematic/quantitative hedge fund (grahamcapital.com operates dedicated 'Quantitative Strategies' and 'Quantitative' pages; Risk.net named it Quant Investment Firm of the Year), so the listed employer genuinely operates in quantitative finance. Identity is supported: the resume is on his own personal site (erikbern.com), corroborated by his LinkedIn (linkedin.com/in/erikbern) and GitHub. His earlier D. E. Shaw & Co. 'Intern Developer' role (Jul-Sep 2007) appears only in the LinkedIn/index-derived enrichment and is NOT corroborated by a primary source, so it is retained as index-only and not counted toward the CONFIRMED verdict; the CONFIRMED verdict rests on the Graham Capital role. Good contacts: erikbern.com/resume.html 21f32.Sep 2011
  • Own resume states he led one half of a team managing a $700M+ portfolio at Graham Capital, implementing HFT strategies in C++ and C# for commodities and currency futures — a genuine quantitative trading role.Mar 2011
  • Created open-source Luigi (workflow engine, 13k+ stars) and Annoy (approximate nearest neighbors) at Spotify; both stayed in wide use (Capital One, Foursquare, Uber, Instacart).Oct 2012
  • Represented Sweden at the IOI in Kenosha, WI in 2003 and placed 8th (gold medal); resume also lists IMO participation.Aug 2003
  • Self-published contact resume lists email [contact omitted], phone 917-940-8790, NYC address.2021
  • Grew Better.com's engineering team from 2 to 300 and personally conducted an estimated 2,000+ interviews.Jan 2021
  • Awarded three patents on recommender systems during his Spotify tenure.2013
  • Master's thesis "Implementing a Scalable Music Recommender System" was awarded best Master's Thesis in Sweden by Naturvetarna.Mar 2009
  • LinkedIn-derived enrichment lists a D. E. Shaw & Co. 'Intern Developer' role (Jul–Sep 2007) that his own resume does not mention; index-only, uncorroborated by a primary source.Jul 2007
  • Resume lists a Google software engineering internship in Zurich (Sep 2006 – Apr 2007) on embedded systems and large-scale data extraction for Google Maps, taken as time off from university.Sep 2006
  • Co-organizer of the NYC Machine Learning meetup (~11,000 members).2018

Competition record

Sweden · IOI

Related collections

Explore the collection