Technical Lead, Multimodal Transformers

Kodiak AI is an active public company providing AI-powered autonomous ground-vehicle technology, principally through the Kodiak Driver for commercial trucking, industrial trucking, and defense applications.

Mountain View, United States
About Kodiak AI, Inc.

Kodiak Driver combines AI-powered software and modular, vehicle-agnostic hardware to enable driverless vehicle operations. Kodiak commercializes the technology through a Driver-as-a-Service model and serves trucking, industrial, and defense customers.

View jobs by Kodiak AI, Inc.

Skills

Candidate Availability

Required and preferred rules are kept separate and reflect the wording in the original posting.

About the Role

You will own the multimodal-transformer architecture roadmap and lead work from problem framing through production deployment. You will direct research on attention, token fusion, and multimodal pretraining; review designs and code; mentor engineers; and align model designs with latency and compute budgets.

Requirements

  • PhD with 5+ years of industry experience, or MS/BS with 8+ years, in AI, computer science, or a related field
  • Experience leading multi-engineer technical efforts from research through production
  • Expertise in transformer architectures in multimodal or multi-stream settings
  • Knowledge of cross-attention, token fusion, and modality-alignment techniques
  • Experience mentoring engineers
  • Python and PyTorch proficiency
  • Experience with large-scale distributed training and mixed-precision optimization
  • Ability to influence technical direction across teams

Responsibilities

  • Own the architecture roadmap for multimodal transformers
  • Lead work from problem framing and experiment design through implementation and production deployment
  • Drive research on cross-modal attention, token fusion, and multi-stream tokenization
  • Set technical standards through design reviews, code reviews, and architectural decisions
  • Mentor junior and mid-level engineers
  • Define self-supervised and contrastive multimodal pretraining strategies
  • Align model design with system-level latency and compute budgets

Benefits

  • Equity
  • Annual bonuses
  • Medical, dental, and vision plans
  • Medical plan with infertility benefits
  • Legal services
  • Identity and fraud protection
  • Hospital indemnity insurance
  • Accident insurance
  • Critical illness insurance
  • Flexible PTO
  • 10 paid holidays
  • Parental leave
  • Dog-friendly office
  • Free catered lunch
  • Stocked kitchen
  • Free EV charging
  • Long-term disability insurance
  • Short-term disability insurance
  • Life insurance
  • Headspace
  • Calm
  • One Medical
  • Gympass
  • Spring Health
  • Rula
  • 401(k)
  • Commuter benefits
  • FSA
  • Dependent Care FSA
  • HSA
  • Referral bonuses
  • Patent bonuses
Technical Lead, Multimodal Transformers at Kodiak AI, Inc. | JobStash