Justin Selig

Investor

2022 - Current

Investing in enterprise AI, semiconductor, and hard tech companies. Led Seed investments in Augmenta AI (sourced, $12M), Efficient Computer (sourced, $16M).

Advisor

2024

Lead Product Manager, Cerebras SDK

2020 - 2022

Founded, launched, and scaled the Cerebras SDK, serving F500 customers across energy, pharma, life sciences, government, and academia. Acted as a General Manager, building teams across engineering, sales, and marketing. Achieved $MM in revenue and enabled new fundamental HPC research. In 2023, my team was nominated for the Gordon Bell Prize in supercomputing. Oversaw all product lines associated with the SDK through 10+ launches, including: • Developer Tools: simulation, hardware/software debug, performance profiling, visualization. • Software Libraries: primitives for machine learning, BLAS, signal processing, etc. • Programming Languages: Cerebras Software Language (CSL), Python, and C++ host APIs. • Framework Integrations: Domain-Specific Language integrations, Deep Learning frameworks. • Documentation: developer portal, programming guides, videos, manuals. The Cerebras SDK is a general-purpose parallel-computing platform for developing ML and HPC software on the world's fastest AI supercomputer, the Cerebras CS-2.

Product Manager, Machine Learning

2019 - 2020

Visualization and developer tooling.

Member Of Technical Staff

2018 - 2020

Deep Learning Kernels Engineer, Parallel-Distributed Algorithms Started many foundational software projects which are now the basis of Cerebras' core kernel infrastructure. My primary focus was on developing high-performance software (assembly microcode, C, C++, Python) for accelerating deep neural networks on the largest computer chip ever built. Day-to-day entailed devising novel dataflow algorithms, implementing and debugging software for linear algebra and collective communication libraries, performance tuning distributed system software, wrestling with nascent compilers, writing technical specifications, and managing cross-disciplinary software projects. My work focused mainly on NLP applications (Seq2seq, GNMT, Transformer, BERT, GPT). Did some fun stuff too... • Onboarded 50+ engineering new hires and technical writers. • Founded our intern events program (hosting 30+ NCGs and Ph.D. interns with 80%+ conversion). • Mentored interns on my software team. • Managed a remote team of SWE contractors contributing to kernel libraries. • Gave technical talks to the company.

Software Engineer, AI Hardware Acceleration and FPGA Design

2017 - 2017

Developed firmware on a soft FPGA microprocessor core for integrating the first 4k image sensor into a computer vision pipeline for Uber’s fleet of self-driving vehicles. Migrated the camera platform to an industry-standard imaging framework, allowing integration testing and V&V to be moved offsite to a contract manufacturer. As a result, reduced high-volume manufacturing costs by an order of magnitude and engineering man-hours for in-house testing by tens of hours per unit.

President

2016 - 2017

EWH is a team of 30+ undergrad and graduate researchers building low-cost IoT medical devices for developing countries. Responsible for product, engineering, business development, and managing a management team. Projects included: - Infant jaundice therapy device. - Mycotoxin filtering table. - Water quality monitoring system with remote operation.

Manager, VP, Electrical and Software Teams

2015 - 2016

Hybrid Product Manager, Engineering Manager, Program Manager, and Engineering Lead role. Projects included: - Portable low-cost vital signs monitor and base station with wireless connectivity. - Smart ECG device.

Software Engineer

2013 - 2015

Avionics Hardware, Embedded Software Engineer

2016 - 2016

Worked in the Avionics RIO (Remote I/O) group on multiple projects focused on the design and delivery of embedded Linux software for the Falcon 9 rocket and Dragon Crew space vehicle. Focused primarily on the telemetry and vehicle flight computers. Projects ranged from firmware driver development, to bit-level error detection in FPGA PCIe links, to EMI/EMC testing of rocket engine PCBAs.

Modeling and Simulation Software Engineer

2015 - 2015

Developed high-fidelity flight simulation tools for real-time image generation in pilot training. Created a framework to replace the standard flight computer UI for prototyping and testing new FMC features (C++, Java).

Backend Software Engineer, Tech Team Leader

2014 - 2015

Built an e-commerce platform focusing mainly on back-end web development (Python, Django, Jekyll). Optimized the LCL marketing website front-end for mobile (HTML, CSS, JQuery). Subsequently served on the LCL management team as the Technical Lead.

Research Assistant

2014 - 2014

Python programmer for various projects in NLP, web scraping, etc. Goal was to learn how to increase civic engagement by creating a sentiment analysis engine for comments on regulationroom.org.

Externship

2014 - 2014