Member of Technical Staff, Multimodal

Boson AI; Santa Clara (CA); On-site; USD 100,000 - 160,000; Full time

Job summary

Boson AI, an innovative startup in Santa Clara, is seeking full-time research scientists and engineers to advance generative multimodal models. In this role, you'll design model architectures and loss objectives while working with diverse datasets across various media formats. Ideal candidates possess a strong machine learning background and are passionate about contributing to cutting-edge AI technology.

Qualifications

• Experience in writing clean and efficient code.
• Participation in at least one research project related to multimodality learning.
• Strong coding and deep learning skills.

Responsibilities

• Design model architectures and loss objectives for multimodal data.
• Build diverse datasets for multimodality learning.
• Develop evaluation pipelines for various generative outputs.

Skills

Machine Learning

Deep Learning

Multimodal Learning

Data Collection

Data Processing

Education

Master’s or Doctoral degree in computer science or equivalent

Tools

PyTorch

JAX

Job description

Boson AI is an early-stage startup building large language tools for everyone to use. Our founders (Alex Smola, Mu Li), and a team of scientists and engineers in Deep Learning, Optimization, NLP, AutoML, and Statistics are working on high-quality generative AI models for language and beyond.

We are seeking research scientists and engineers to join our team full-time in our Santa Clara office. Your role will involve designing model architectures, proposing new loss objectives, and advancing generative multimodal models. The ideal candidate has a strong machine learning background and is motivated to develop state-of-the-art models towards AGI.

We encourage you to apply even if you do not meet every qualification. If you are motivated to learn and contribute to foundation models, we would love to chat.

Responsibilities

• Design model architectures and loss objectives for multimodal data (images, video, text, speech, audio)
• Build diverse datasets for multimodality learning, including data collection and processing
• Develop evaluation pipelines for various generative outputs

Qualifications

• Experience in writing clean and efficient code; Master’s or Doctoral degree in computer science or equivalent
• Proficiency in deep learning frameworks like PyTorch or JAX
• Participation in at least one research project related to multimodality learning

Preferred Qualifications

• Experience in multimodal joint embedding, text-to-image/video generation, etc.
• Experience in document understanding (layout analysis, data extraction, OCR)
• Experience in audio processing tasks; active GitHub contributions are a plus
• Experience handling data at billions-scale