Software Engineer - San Francisco, United States - Twelve Labs

Twelve Labs San Francisco, United States

4 weeks ago

Description

Who we are
We're a fast-moving, diverse team pushing the frontiers of artificial intelligence. At Twelve Labs, our mission is to help developers build programs that can see, listen, and understand the world as we do by bringing the world's most powerful video understanding infrastructure to market. As a part of achieving this mission, we are building foundation AI models that can accurately and instantly search exact moments within petabytes of video archives, generate coherent text summaries of videos, perform prompt-based video generation, and many more. The Twelve Labs platform provides access to its Large Visual Language Models (VLMs) through a suite of APIs that are trained on massive video datasets and learn to understand the meaning and context behind the visuals, conversations, and sounds within videos.
Twelve Labs recently raised $17M in seed funding, recognized as one of CB Insights' AI 100 companies within a year of its founding, and secured a massive compute resource through partnering with Oracle. We are hyper focused on delivering the Twelve Labs platform to our customers so they can build video understanding into their products and power dream features they could have only imagined.
Part of the pathway to our rapid growth has been paved by the outstanding group of people united by the company's mission. Beyond prominent venture capital firms such as Index Ventures and Radical Ventures, the Twelve Labs mission is backed by category building luminaries like Fei-Fei Li (Stanford HAI), Silvio Savarese (Salesforce), Oren Etzioni (AI2), Alexandr Wang (Scale), Lukas Biewald (W&B), Jack Conte (Patreon) and more.
We are committed to creating a diverse and inclusive work environment where our team members can bring their full selves to work, bring out their potential, and most importantly, thrive together. We welcome kind, brilliant, and open minded people from all walks of life to our team. If joining this mission speaks to you, we encourage you to apply
About the Role:
As a Software Engineer - Machine Learning at Twelve Labs, you will be a vital member of the ML Deployment & Operations Team. Your primary role is to build and deploy the machine learning pipeline in our ML Infrastructure while using the foundation models provided by the ML Modeling & Research Team. You will ensure seamless deployment of models end-to-end and implement best MLOps practices to automate the integration, deployment, and training process. A critical KPI for this role is minimizing the time from model training to deployment on our machine learning infrastructure and serving the model as efficiently as possible in terms of latency and throughput. We're looking for someone who is excited to collaborate across ML Infrastructure, ML Modeling, and Data Team.

You will:

Be responsible for Model Serving and ModelOps: manage model-related metadata (using the model registry), implement hardware-accelerated optimization for each model engine, and containerize models for efficient serving.
Construct an ML pipeline that proficiently serves the trained foundation models in our ML Infrastructure.
Implement best model validation practices by conducting automatic evaluation benchmarking and performing output comparisons.
Develop an automatic training/finetune pipeline that includes rigorous data and model validation against the baseline model.

You may be a good fit if you have:

5+ years of software development experience, including experience in deploying machine learning models
3+ years of experience in building and deploying an end-to-end machine learning pipeline, or equivalent
Have experience in establishing and maintaining secure software and system development environments
Have experience designing control and sandboxing systems for AI research
Willingness to learn the emerging AI technology and a practical mindset on productization
Have a black-box level of understanding in Transformer-based neural network
Experience in system development in model serving and inference

Desired Experience:

Experience with MLOps for managing the entire machine learning lifecycle, including model registry and versioning functionalities
Experience with hardware-accelerated optimization techniques
Experience in identifying and mitigating pain-points in ML research & modeling processes
ML research experience would be helpful, as this role requires interchangeable effort on both research side and software side

Relevant Tech Stack:

Language: Python, C++, CUDA
ML / Platform: PyTorch, Docker, Kubernetes
ML Demo page: Gradio, Streamlit
MLOps: MLFlow, Weights and Biases
Automation: Airflow, Kubeflow
Model serving: Triton, FasterTransformer

Interview and Onboarding Process
Recruiter Phone Screen -> Hiring Manager Call -> Technical Interview and/or Take Home Assignment -> Culture Interview -> Reference Checks
We're also excited to share that we'll do global onboarding in Seoul for all new hires (company-sponsored travel).
Even If there are a few checkboxes that aren't ticked through your prior experience, we still encourage you to apply If you are a 0-to-1 achiever, a ferocious learner, and a kind and fun team player who motivates others, you will find a home at Twelve Labs.
We welcome applicants from all walks of life and are committed to equal opportunity employment. We cherish and celebrate diversity not just because it is the right thing to do, but because it makes our company much stronger.

Benefits and Perks
An open and inclusive culture and work environment. Work closely with a collaborative, mission-driven team on cutting-edge AI technology. Full health, dental, and vision benefits Extremely flexible PTO and parental leave policy. Office closed the week of Christmas and New Years. Remote-flexible, offices in San Francisco and Seoul and coworking stipend VISA support (such as H1B and OPT transfer for US employees)