Asks for

The ad says“LLM platforms - Anthropic Claude (primary), Microsoft Copilot, Google Gemini, OpenAI; hands-on with APIs, agent frameworks, and production deployment across at least one”
The ad says“Cloud platforms - AWS, Azure, or GCP; deployment, containers, infrastructure basics”
The ad says“Cloud platforms - AWS, Azure, or GCP; deployment, containers, infrastructure basics”
The ad says“Implement advanced LLM patterns where required: retrieval-augmented generation (RAG), chain-of-thought reasoning, tool use, multi-agent orchestration, and structured output validation”
The ad says“Claude Skills framework and Cowork - structured instruction design, output validation, skill packaging”
The ad says“Cloud platforms - AWS, Azure, or GCP; deployment, containers, infrastructure basics”
The ad says“LLM platforms - Anthropic Claude (primary), Microsoft Copilot, Google Gemini, OpenAI; hands-on with APIs, agent frameworks, and production deployment across at least one”
The ad says“REST and GraphQL APIs - authentication, error handling, rate limiting, webhooks”
The ad says“LLM evaluation and observability - LangSmith, Weights & Biases, or equivalent”
The ad says“LLM platforms - Anthropic Claude (primary), Microsoft Copilot, Google Gemini, OpenAI; hands-on with APIs, agent frameworks, and production deployment across at least one”
The ad says“MCP (Model Context Protocol) - connector design, enterprise system integration, authentication flows”
The ad says“LLM platforms - Anthropic Claude (primary), Microsoft Copilot, Google Gemini, OpenAI; hands-on with APIs, agent frameworks, and production deployment across at least one”
The ad says“Implement advanced LLM patterns where required: retrieval-augmented generation (RAG), chain-of-thought reasoning, tool use, multi-agent orchestration, and structured output validation”
The ad says“Python and/or Node.js - production-grade, testable, maintainable code”
The ad says“LLM platforms - Anthropic Claude (primary), Microsoft Copilot, Google Gemini, OpenAI; hands-on with APIs, agent frameworks, and production deployment across at least one”
The ad says“Python and/or Node.js - production-grade, testable, maintainable code”
The ad says“REST and GraphQL APIs - authentication, error handling, rate limiting, webhooks”
The ad says“Implement advanced LLM patterns where required: retrieval-augmented generation (RAG), chain-of-thought reasoning, tool use, multi-agent orchestration, and structured output validation”
The ad says“LLM evaluation and observability - LangSmith, Weights & Biases, or equivalent”

Greenlight helps organizations solve complex business challenges through intelligent automation, agentic AI, and custom technology solutions.

Our teams work directly with clients to understand their operations, identify opportunities, and rapidly build solutions that create measurable business value. We combine deep consulting expertise with hands-on engineering to bridge the gap between strategy and execution.

We’re building a future where consultants and engineers work alongside AI to deliver faster outcomes, stronger businesses, and transformative customer experiences. Anthropic’s Claude is embedded across how we design, build, and validate solutions - and this role is at the technical frontier of that capability.

What makes a star at Greenlight?

· Builder's instinct, architect's discipline

· The most technically credible person in the room

· AI-native at depth - you know how LLMs fail and build around it

· Client-embedded, not arm's length

· Commercially aware - scope, economics, outcomes

The Role

The Forward Deployed Engineer is Greenlight’s most senior technical delivery role. You design, build, and operationalize AI agent workflows and intelligent automation solutions - embedded directly in client environments, working alongside their teams to deliver production-grade solutions that create measurable business outcomes.

This is not a back-office engineering role. You will be in client discovery sessions, presenting architectures to IT leadership, deploying AI agents against real enterprise systems, and building the reusable skills and accelerators that make Greenlight faster and more differentiated on every subsequent engagement. You are expected to be the most technically credible person in the room - combining deep AI fluency with the consulting presence of a senior practitioner.

You work at the frontier of agentic AI - designing and building Claude-powered solutions, connecting AI agents to enterprise systems through MCP integrations, and producing the technical documentation that the delivery team and the client can both stand behind. The platform is Anthropic Claude. The problem space is complex, regulated, and high-stakes. The bar is production-grade.

At a Glance

Reports To	AI Practice Lead
Works Closely With	Pre-Sales SE, Automation Business Consultant, AI Delivery Engagement Manager
Client Interaction	Yes — C-suite, IT architects, operations leaders, technical teams
Travel Requirement	Regular client travel required — discovery, workshops, POC delivery, go-live
Platform Focus	Anthropic Claude (Cowork + Skills), MCP, Python / Node.js, REST APIs
Seniority	Intermediate 3-5 year’s experience
Location	Onshore Canada — Toronto preferred
Engagement Type	Hybrid — onsite client-facing with remote delivery phases

What You’ll Do

Solution Architecture & Technical Discovery

Lead technical discovery workshops with C-suite, IT architects, and operations leaders - identifying high-value AI automation opportunities and translating complex operational workflows into executable solution architectures
Design end-to-end AI agent architectures using Anthropic’s Claude platform (Cowork and Skills framework), aligned to client infrastructure, security requirements, and compliance constraints
Translate business requirements into agent-based automation blueprints: what the AI owns, what stays human, how exceptions route, and how the two coordinate
Assess the right solution architecture for each use case - knowing when a Claude agent is the answer, when a procode integration is cleaner, and when a simpler rules-based approach is more appropriate than AI
Present technical architectures to executive and non-technical audiences with clarity, confidence, and the credibility that comes from having built things like this before

AI Agent Build & Delivery

Engineer production-ready AI agent workflows using Claude Cowork and the Skills framework - built to a standard that holds up in regulated, enterprise production environments
Develop and configure Claude Skills: structured instruction sets, templates, validation rules, and reference documents that encode client-specific processes and quality standards into repeatable AI agent behaviour
Build and maintain MCP connectors that integrate AI agents with enterprise systems - CRMs, ERPs, document management platforms, core banking systems, and custom APIs - handling authentication, data mapping, error handling, and audit requirements
Implement advanced LLM patterns where required: retrieval-augmented generation (RAG), chain-of-thought reasoning, tool use, multi-agent orchestration, and structured output validation
Conduct output validation, performance tuning, and safety testing of deployed AI agents before client go-live - quality is non-negotiable when outputs go to regulators or clients

Procode & Integration Engineering

Write production-quality Python or Node.js to extend platform capability - data pipelines, transformation logic, webhook handlers, API wrappers, and back-end components that sit outside the automation itself
Build, test, and maintain integrations with enterprise platforms (Salesforce, Jira, DocuSign, SharePoint, ServiceNow, and others) using REST, GraphQL, OAuth 2.0, and enterprise authentication patterns
Design and implement monitoring, logging, and observability into deployed solutions so operations teams can diagnose issues without requiring an engineer in the room
Apply cloud platform fundamentals (AWS, Azure, or GCP) to deployment, scaling, and infrastructure decisions where the engagement requires it

Accelerators & Practice Development

Build and continuously refine reusable Skills libraries, automation templates, and reference architectures that make every subsequent Greenlight engagement faster and more defensible
Contribute to pre-sales activities: technical demos, POC builds, RFP responses, and solution estimations - your hands-on credibility is a direct input to Greenlight’s win rate
Produce high-quality technical documentation - solution design documents, architecture diagrams, runbooks, and post-deployment reports - written to a standard the client’s IT team can maintain
Identify account expansion opportunities within active engagements - additional use cases, adjacent processes, and follow-on automation candidates - and surface them to the AI Delivery Engagement Manager
Mentor Automation Engineers and client technical teams on AI agent patterns, LLM integration best practices, and platform capabilities

Client Engagement & Consulting Delivery

Operate with high autonomy inside client environments - navigating ambiguity, defining structure where none exists, and delivering without requiring hand-holding from the practice lead
Manage scope, timelines, and deliverables across concurrent client engagements - you know when to flag a risk and when to absorb it
Build trusted-advisor relationships with client technical teams - the goal is to be the person they call when they have an AI or automation problem, not just the vendor who delivered the last project
Deliver working solutions within 2–4 week sprint cycles - you build and show, you don’t slide and tell

What We’re Looking For

Skills & Competencies

Technical depth with consulting presence - credible with a CTO and a COO in the same meeting
AI-native - you understand how LLMs behave under pressure, where they fail, and how to build reliable systems around them
Builder's bias - working software over slide decks, every time
High autonomy - you define structure where none exists without needing direction
Commercially aware - scope, delivery economics, and what it costs when technical decisions go sideways
Clear communicator - architecture decisions and trade-offs explained in language both business and IT can act on

Technical Familiarity

LLM platforms - Anthropic Claude (primary), Microsoft Copilot, Google Gemini, OpenAI; hands-on with APIs, agent frameworks, and production deployment across at least one
Claude Skills framework and Cowork - structured instruction design, output validation, skill packaging
MCP (Model Context Protocol) - connector design, enterprise system integration, authentication flows
Python and/or Node.js - production-grade, testable, maintainable code
REST and GraphQL APIs - authentication, error handling, rate limiting, webhooks
Cloud platforms - AWS, Azure, or GCP; deployment, containers, infrastructure basics
LLM evaluation and observability - LangSmith, Weights & Biases, or equivalent

What Great Looks Like in This Role

The standard we are hiring to:

The best FDEs we have worked with walk into a client environment and within a week they know more about the client’s process than the client’s own team has documented. They build a working POC in a few weeks that makes the client stop asking whether it’s possible and start asking when it can go to production. They write code that the next engineer can understand without a two-hour briefing. They know when the LLM is going to fail before it does, and they’ve already built the fallback. And they leave every engagement with a client technical team that is more capable than when they arrived - because that is what makes clients come back.

AI in Our Hiring Process

We use AI tools to support parts of our recruitment - organizing applications and flagging relevant experience. These tools inform our process, they don’t drive it. Every hiring decision is made by our team, full stop.

Compensation

The expected salary range for this role is $90,000 – $130,000 CAD annually, based on experience and qualifications.