Signals and Threads

In Young Cho, Co-founder and CEO of DeepMind, discusses her unexpected transition from medicine to finance through a trading internship. She delves into the integration of trading, research, and software engineering. The conversation highlights the challenges of machine learning in chaotic environments and the evolution from basic models to deep neural networks. Cho also addresses the importance of robust data practices and explores the complexities of using Python notebooks in financial research.

Sylvain Gugger, a machine learning engineer at Jane Street and co-author of "Deep Learning for Coders," shares his fascinating transition from teaching math to ML. He delves into optimizing learning rate strategies and the nuances of working with PyTorch. The conversation touches on the importance of reproducibility in training models as well as the challenges of inference in trading, emphasizing low latency with uniquely shaped market data. Sylvain also highlights Hugging Face's role in making ML tools more accessible, enhancing collaboration within the field.

Liora Friedberg, a Production Engineer at Jane Street with expertise in economics and computer science, discusses the unique blend of puzzle-solving and software engineering in high-stakes environments. She shares insights on training methods, including tabletop simulations and hands-on exercises, to prepare engineers for the complexities of live system support. Liora also highlights the importance of fostering a blame-free culture, effective monitoring systems, and the art of postmortems to encourage learning and improve operational support.

Erin Murphy, Jane Street’s first UX designer and former NASA designer, discusses challenges of user-centered design in expert environments. They delve into bringing web interfaces to power users, beauty in design, Jane Street's UX candidate criteria, and helping engineers understand user needs.

Andrew Hunter, a performance engineer who makes code really fast, discusses optimizing systems at hyperscale and the challenges of optimizing trading systems. He shares his favorite profiling techniques, tools for visualizing traces, and the unique challenges of optimizing OCaml versus C++. They also touch on the joys of musical theater and how to pass an interview when sleep-deprived.

Peter Bogart-Johnson, a poet and program manager at Jane Street, discusses the challenge of gaining trust as an outsider, strategies for teaching new ways of working, and the importance of listening as a PM. They also touch on paying down technical debt, qualities Jane Street looks for in PM candidates, and coordinating teams during times of change.

Richard Eisenberg is one of the core maintainers of Haskell. He recently joined Jane Street’s Tools and Compilers team, where he hacks on the OCaml compiler. He and Ron discuss the powerful language feature that got him into PL design in the first place—dependent types—and its role in a world where AIs can (somewhat) competently write your code for you. They also discuss the differences between Haskell and OCaml; the perils of trying to make a language that works for everybody; and how best a company like Jane Street can collaborate with the open source community.You can find the transcript for this episode  on our website.Some links to topics that came up in the discussion:Dependent typesGHCUnboxed types in OCamlLanguage extensions in Haskell

Ella Ehrlich has been a developer at Jane Street for close to a decade. During much of that time, she’s worked on Gord, one of Jane Street’s oldest and most critical systems, which is responsible for normalizing and distributing the firm’s trading data. Ella and Ron talk about how to grow and modernize a legacy system without compromising uptime, why game developers are the “musicians of software,” and some of the work Jane Street has done to try to hire a more diverse set of software engineers.You can find the transcript for this episode  on our website.Some links to topics that came up in the discussion:EG, The League of Legends team that Ella is a huge fan of.Apache Kafka, the message bus that Gord migrated to.Some of the various sources of symbology you have to deal with when normalizing trading data. (Really, there are too many sources to list here!)A list of Jane Street’s recruiting Programs and Events, including INSIGHT, which focuses on women, and IN FOCUS, which focuses on historically underrepresented ethnic or racial minorities.

Doug Patti is a developer in Jane Street’s Client-Facing Tech team, where he works on a system called Concord that undergirds Jane Street’s client offerings. In this episode, Doug and Ron discuss how Concord, which has state-machine replication as its core abstraction, helps Jane Street achieve the reliability, scalability, and speed that the client business demands. They’ll also discuss Doug’s involvement in building a successor system called Aria, which is designed to deliver those same benefits to a much wider audience.You can find the transcript for this episode  on our website.Some links to topics that came up in the discussion:Jane Street’s client-facing trading platformsA Signals and Threads episode on market data and multicast which discusses some of the history of state-machine replication in the markets.The FIX protocolUDP multicastReliable multicastKafka

Stephen Dolan works on Jane Street’s Tools and Compilers team where he focuses on the OCaml compiler. In this episode, Stephen and Ron take a trip down memory lane, discussing how to manage computer memory efficiently and safely. They consider trade-offs between reference counting and garbage collection, the surprising gains achieved by prefetching, and how new language features like local allocation and unboxed types could give OCaml users more control over their memory.You can find the transcript for this episode  on our website.Some links to topics that came up in the discussion:Stephen’s command-line JSON processor, jqStephen’s Cambridge dissertation, “Algebraic Subtyping”, and a protoype implementation of mlsub, a language based on those ideas.A post from Stephen on how to benchmark different memory allocators.A Jane Street tech talk on “Unboxed Types for OCaml”, and an RFC in the OCaml RFC repo.A paper from Stephen and KC Sivaramakrishnan called “Bounding Data Races in Space and Time”, which is all about a new and better memory model for Multicore OCaml.Another paper describing the design of OCaml’s multicore GC.The Rust RFC for Higher-ranked trait bounds.