Rust Engineer - Real-Time Inference & Architecture (Remote)

Bot Jobs; Myrtle Point (OR); Remote; USD 86,000 - 138,000; Full time

Benefits offered by this job

Technical ownership

R&D freedom

Competitive compensation

Compute infrastructure

Flexible remote

Job summary

Deepslate is seeking a Rust Engineer to join a fully remote team working on real-time inference engines and the Rust-based model stack. You will own critical components, design architecture, and contribute to production-ready systems.

The role emphasizes writing high-quality Rust code, building scalable architectures, and optimizing performance for low-latency real-time responses in conversational models.

Qualifications

• Strong hands-on experience with Rust, including ownership, borrowing, composition, and idiomatic Rust design patterns.
• Solid experience with async Rust and structuring concurrent applications.
• Experience building and working with microservice architectures.
• A true startup mindset: comfortable navigating early-stage codebases and limited docs.
• Drive and vision: actively shape the product, architect solutions, and own projects.

Responsibilities

• Build the inference engine and core components of the model stack in Rust from architecture to production.
• Design the architecture for component communication and extensibility.
• Develop real-time async Rust systems with low latency for streaming audio.
• Identify bottlenecks and optimize latency, throughput, and resource usage.
• Write production-ready code and improve tooling and documentation.

Skills

Rust

async Rust

microservice architectures

startup mindset

ownership & architecture

Tools

bindgen

Job description

Deepslate is seeking a Rust Engineer to join a fully remote team working on real-time inference engines and the Rust-based model stack. You will own critical components, design architecture, and contribute to production-ready systems.

The role emphasizes writing high-quality Rust code, building scalable architectures, and optimizing performance for low-latency real-time responses in conversational models.