Software Engineer, Adept

Distributed Training

$175-350k

+ Stock options

Senior and Expert level

San Francisco Bay Area

Machine learning research and product lab

Job no longer available

Machine learning research and product lab

21-100 employees

B2BArtificial IntelligenceRoboticsMachine Learning

Job no longer available

$175-350k

+ Stock options

Senior and Expert level

San Francisco Bay Area

21-100 employees

B2BArtificial IntelligenceRoboticsMachine Learning

Company mission

To solve open problems in AI in order to train a neural network that knows how to use every piece of software we use in our lives.

Job

Company

Role

Who you are

We value curious engineers who can engage with new problems and get things done at a startup
8+ years of experience as a software engineer
Expert understanding of distributed training concepts and tools, e.g., torch.distributed, NCCL, MPI, etc
Experience with GPU cluster hardware, performance, interconnect, etc
Track record at fast-growing companies or startups
Demonstrated end-to-end ownership and self-direction
Comfort with moving fast and learning by doing
Excellent communication and collaboration skills, both verbal and written

What the job involves

Adept is building a new class of multimodal AI models designed specifically for digital agents
Our model iteration speed depends on training performance while using some of the largest GPU clusters around
Engineers with a background in distributed deep learning (DL) training can help Adept iterate faster
Areas of focus on the Infrastructure team include:
Compute - building and managing large GPU clusters training SOTA models
Optimization - improving the DL training utilization and reliability of those clusters
Research - working directly with researchers to align model architecture with training performance

Our take

Adept is developing AI models that don't just respond to text commands - like a chatbot - but actually turn those commands into actions. The company's generative AI has the potential to help users carry out tasks on their computers, from browsing the internet to navigating enterprise software tools.

The company's model, ACT-1 is designed and trained to take action on users' computers in response to their natural language commands. While such technology is still in early development, with the speed of innovation in the AI space, CEO David Luan believes the Adept advanced AI capabilities are only a couple of years away, not decades.

Fairly new to a fiercely competitive industry, Adept was only founded in 2022 but that hasn't stopped it from gaining the support of some of the largest companies in technology, including Nvidia, Microsoft, Workday, and Atlassian. With several rounds of funding under its belt, Adept is equipped to accelerate the training of its AI models and reach its goal of creating the next-generation generative AI model.

Steph

Company Specialist at Welcome to the Jungle

Company

Funding (2 rounds)

Mar 2023

$350m

SERIES B

Apr 2022

$65m

SERIES A

Total funding: $415m

Company benefits

Comprehensive health insurance coverage - 100% for employees
Dental and vision insurance
Unlimited vacation time for exempt employees
4 remote weeks per year - work from anywhere
Stock options
Daily meals for those in our comfortable SF office
Commuter benefits

Company HQ

SoMa, San Francisco, CA

Leadership

David Luan

(CEO & Co-Founder)

Previously worked as the Director of AI at Axon, the Director of Engineering at OpenAI, and as a Director at Google Research.

Fred Bertsch

(Co-Founder)

Experienced Software Engineer having previously worked at Google.

Salary benchmarks

We don't have enough data yet to provide salary benchmarks for this role.

Submit your salary to help other candidates with crowdsourced salary estimates.

Share this job