Founding ML Engineer

Job Description

About the Role:

As a Founding ML Engineer means working on cutting-edge technology to create a digital humanoid coworker that can operate alongside humans using the same tools. This role involves designing and training models for multi-step workflows, experimenting with ML models, and building low latency inference infrastructure. You will be part of the founding team, collaborating closely with customers and contributing to the growth and success of the company.

Role Responsibilities:

Design and train models to run multi-step workflows by operating software directly (via keyboard and mouse) and calling APIs.
Experiment with and fine-tune existing ML models to find the right balance between size, accuracy, and speed.
Design benchmarks to improve our understanding of data and model performance.
Build low latency inference infrastructure for speech, language, and reasoning models.
Collaborate closely with customers as part of the founding team.

Required Skills:

5+ years of professional experience.
Experience with LAM modelsagents that can reason and perform actions to accomplish tasks.
Training, fine-tuning, and prompt engineering of generative models and LLMs.
Experience in early-stage and fast-growing companies.
Used and tuned ASR and TTS models.
Experience with vector databases, embedding models, and building real-world RAG pipelines.
Familiarity with state-of-the-art LLMs and their strengths/weaknesses.
Experience in 0-to-1 work on ML models and infrastructure.
High level of autonomy and self-direction.
3+ years of experience working on deep learning.
Experience in building, deploying, and running ML infrastructure.
Strong Python programming skills and expertise in ML frameworks like PyTorch and TensorFlow.
Please let me know if you are interested.

Work in United States

Employment Options

Base Salary

160,000 - 220,000 USD

Required Academic Degree

Professional Experience

Skills

• Python
• PyTorch
• TensorFlow
• ML Infrastructure
• ML Models
• LAM Models
• ML Frameworks
• Deep Learning