Full-Time

GPU Kernel Engineer

Updated on 8/20/2026

TypeSafe AI

TypeSafe AI

Reliable AI models for embedding into software

Compensation Overview

$150k - $250k/yr

+ Equity

H1B Sponsorship Available

San Francisco, CA, USA

In Person

Category
Software Engineering (1)
Required Skills
Graphics Processing Unit (GPU)
CUDA
Requirements
  • Deep CUDA and GPU kernel expertise with a track record of real performance improvements.
  • Experience building and optimizing inference and training kernels.
  • Hands-on large language model training experience.
  • Ability to reason from first principles about performance, memory, and parallelism.
Responsibilities
  • Write, optimize, and maintain high-performance GPU kernels, including kernels written in CUDA or the CuTe domain-specific language, for training and inference.
  • Profile end-to-end performance and eliminate bottlenecks across the stack.
  • Partner with research and platform engineers to maximize hardware throughput and minimize latency.

TypeSafe AI is an early-stage company developing a new class of artificial intelligence models designed to embed reliable, composable intelligence into traditional software systems. The company frames its work around "Reliable Artificial Intelligence," building models intended to act as dependable components that engineers integrate into existing applications rather than as standalone tools. Its models are positioned to bring predictable, composable behavior to enterprise-scale software, letting AI-driven decisions be wired into conventional codebases and workflows. What sets TypeSafe AI apart is its emphasis on reliability and composability as first-class properties, treating AI as an embeddable software primitive rather than an add-on; the company is currently pre-launch, gathering interest through a waitlist and hiring its team. The goal is to make AI a trustworthy, integrable building block for the software businesses already run.

Company Size

N/A

Company Stage

N/A

Total Funding

N/A

Headquarters

N/A

Founded

2024

Simplify Jobs

Simplify's Take

What believers are saying

  • Almeida said TypeSafe AI raised over $35 million by March 2026.
  • The company started opening access to top-tier engineering teams in January 2026.
  • Recent 2026 talks suggest strong thought-leadership and founder credibility in frontier AI.

What critics are saying

  • TypeSafe AI has no visible customer list, making revenue quality impossible to verify.
  • A waitlist-only funnel signals early product-market fit risk into 2027.
  • If reliability claims fail, enterprise buyers will choose OpenAI, Anthropic, or open-source stacks.

What makes TypeSafe AI unique

  • Diogo Almeida says TypeSafe AI builds models for no-human-in-the-loop automation, not chat.
  • TypeSafe AI positions reliability, latency, and cost as the core product wedge.
  • The company explicitly targets AI embedded inside real systems, not assistant workflows.

Help us improve and share your feedback! Did you find this helpful?

Benefits

Health Insurance

Meal Benefits