Research Software Engineer: Real-Time & LLM Systems

Socket.dev; Montreal (administrative region); On-site; CAD 120,000 - 180,000; Full time

Benefits offered by this job

Mentorship

Postgrad support

Language tutoring

Job summary

DRW Montreal is seeking a software engineer who builds data-intensive, real-time systems. You will design and implement production-grade agentic tooling, manage end-to-end performance, and drive signal-based improvements across distributed components.

Strong bias for rapid iteration and measurable outcomes is essential. You will work at the intersection of research and infrastructure, leveraging LLMs to extract, classify, triage, and accelerate discovery while maintaining rigorous reliability

Qualifications

• 3–5 years building real-time or data-intensive systems with strong trajectory.
• Deep knowledge of LLMs, contexts, tool use, and production harness design.
• Strong Python/Go/Rust fundamentals with deep systems experience.
• Deep knowledge of network protocols: TCP/UDP/IP, DNS, HTTP(S), WebSocket, QUIC.
• Rapid POCs to production and ability to explain results to diverse audiences.
• Bachelor's or Master's in CS, Data Science, Mathematics, or related field.

Responsibilities

• Expand event-driven strategy across feeds, news, social data, and telemetry.
• Run rapid Python analyses to assess signal quality and drive system improvements.
• Build agentic harnesses with tools, context mgmt, evals, guardrails in production.
• Apply LLMs where they add value and avoid them where unnecessary.
• Own full-stack performance from network edge to in-memory stores with SLOs.

Skills

Real-time systems

Data-intensive systems

LLMs integration

Python

Go

Rust

Network programming

Rapid POCs to production

Computer Science fundamentals

Education

Bachelor's or Master's in CS/Data Science/Math

Tools

NGINX/Envoy

Load balancing

Docker

Kubernetes

Job description

DRW Montreal is seeking a software engineer who builds data-intensive, real-time systems. You will design and implement production-grade agentic tooling, manage end-to-end performance, and drive signal-based improvements across distributed components.

Strong bias for rapid iteration and measurable outcomes is essential. You will work at the intersection of research and infrastructure, leveraging LLMs to extract, classify, triage, and accelerate discovery while maintaining rigorous reliability