Senior Software Developer, AI Solutions

System Soft Technologies; Quebec; On-site; CAD 110,000 - 170,000; Full time

Benefits offered by this job

Production agentic systems experience

End‑to‑end ownership of delivery

GenAI tooling exposure (LangChain, LLa

Cloud deployments (AWS, Azure, GCP)

Job summary

System Soft Technologies is seeking a senior software engineer for the AI Solutions team. This hands‑on role focuses on building production agentic systems for customers and clients, using agentic building blocks such as context, memory, guardrails, and orchestration. You’ll shape architecture, guide agents, and validate results in a fast‑paced SDLC.

This is a software development role, not ML research, requiring ownership of end‑to‑end delivery and strong communication of specs and decisions.

Qualifications

• Strong software‑engineering foundation with production software you’ve built, shipped, and owned.
• CS/CE degree, 7+ years’ experience, 5+ in Python.
• Solid architecture and system design – decompose an ambiguous problem, design the solution, guide humans or agents to build it.
• Daily fluent use of agentic coding tools (Claude Code, Codex, Copilot).
• Fluency with the building blocks of agentic AI: context, memory, guardrails, tool calling / MCP, and skills.
• Judgment to validate AI output – hallucinated APIs, risky dependencies, security regressions.
• Clear communication, precise specification writing, and Git proficiency.

Responsibilities

• Build agentic AI solutions for customers and client from the core building blocks: context, memory, guardrails, tool calling / MCP, skills, and orchestration.
• Design the harness, context, and loops that let agents do real development — including agents that build software.
• Use agentic coding tools daily to maximize output, then attack the next bottleneck in the SDLC.
• Steer agents with software‑architecture direction and validate their output — hallucinated APIs, risky dependencies, security regressions.
• Evaluate and adopt emerging AI tooling into Syntax’s practices.
• Own delivery end to end and communicate clearly on specs, decisions, and roadblocks.

Skills

Production software ownership

Software architecture

Agent orchestration

Python programming

Agentic coding tools usage

Education

CS/CE degree

Tools

Claude Code

Codex

Copilot

LangChain

LlamaIndex

CrewAI

GitLab

GitLab CI

Job description

client is looking for a senior software engineer — a builder first — for our AI Solutions team. This is a hands‑on development role, not machine‑learning research: you’ll build production agentic solutions for our customers and for client from the building blocks of agentic systems — context, memory, guardrails, orchestration. You’re past the debate about using AI coding tools — you use them daily to ship more — and you bring the architecture instinct to design systems, guide agents, and validate what they build.

Note: this is a software development role, not ML research or modeling.

Responsibilities:

• Build agentic AI solutions for customers and client from the core building blocks: context, memory, guardrails, tool calling / MCP, skills, and orchestration.
• Design the harness, context, and loops that let agents do real development — including agents that build software.
• Use agentic coding tools daily to maximize output, then attack the next bottleneck in the SDLC.
• Steer agents with software‑architecture direction and validate their output — hallucinated APIs, risky dependencies, security regressions.
• Evaluate and adopt emerging AI tooling into Syntax’s practices.
• Own delivery end to end and communicate clearly on specs, decisions, and roadblocks.

Qualifications:

• Strong software‑engineering foundation with production software you’ve built, shipped, and owned.
• CS/CE degree, 7+ years’ experience, 5+ in Python.
• Solid architecture and system design — decompose an ambiguous problem, design the solution, guide humans or agents to build it.
• Daily, fluent use of agentic coding tools (Claude Code, Codex, Copilot).
• Fluency with the building blocks of agentic AI: context, memory, guardrails, tool calling / MCP, and skills.
• Judgment to validate AI output — hallucinated APIs, risky dependencies, security regressions.
• Clear communication, precise specification writing, and Git proficiency.

Plus:

• Shipped production agentic systems or AI‑powered applications end to end.
• Contributed to MCP servers/tools, eval frameworks, or agent‑orchestration frameworks.
• GenAI frameworks (LangChain, LlamaIndex, CrewAI) and model evaluation; GitLab and GitLab CI.
• Cloud services and deployments (AWS, Azure, GCP).