AI Lead Consultant, AI Practice

AI Lead Consultant

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 are building a future where consultants and engineers work alongside AI to deliver faster outcomes, stronger businesses, and transformative customer experiences. We use Anthropic's Claude as a core delivery tool - and this role sits at the center of how we define what gets built and why it matters.

What makes a star at Greenlight?

Thinks like a consultant, challenges like someone who has seen AI projects fail

Embedded onsite - you live in the client's world until the solution works

AI-fluent without being an engineer - you know what agents can and can't do

Builds the business case before recommending the solution

Iterative by instinct - you refine as you learn, not after the project closes

The Role

The AI Lead Consultant is the analytical and commercial intelligence layer of every Greenlight AI engagement. You are not an engineer - you are the person who ensures the engineer builds the right thing. Working in a two-in-a-box model alongside a Forward Deployed Engineer, you own the what and the why of every automation initiative: why this process, what it should do, how success is measured, and what the business must look like after the AI agent is deployed.

The most common failure mode in enterprise AI is not a technology failure - it is a requirements failure. Engineers build exactly what they are asked to build, and what they are asked to build is often a precise replication of a broken process, wrapped in AI. The result: faster execution of the wrong workflow. This role is the structural safeguard against that outcome.

You will travel to client sites, run discovery workshops, challenge process assumptions, design the business logic that AI agents will execute, build the business case that justifies investment, and produce the requirements that the FDE builds against without ambiguity. You are client-facing, delivery-accountable, and commercially aware. This is not a back-office BA role.

At a Glance

Reports To

Practice Lead

Works Closely With

Forward Deployed Engineer, Pre-Sales SE, AI Delivery Engagement Manager

Client Interaction

Yes - C-suite, operations leaders, process owners, compliance stakeholders

Travel Requirement

Regular client travel required - discovery, workshops, executive readouts

Platform Focus

Anthropic Claude (Cowork + Skills), Claude.ai, AI agent frameworks

Seniority

Mid-to-Senior (3-5 years relevant experience)

Location

Onshore Canada - GTA preferred

Engagement Type

Hybrid - embedded onsite client delivery with remote phases

The Two-in-a-Box Model

Every AI Practice engagement runs with two people who together form a complete delivery unit:

Role

Owns

How They Work Together

AI Lead Consultant ★

The what and why - process discovery, requirements, business case, stakeholder management, AI logic design

You define what gets built and why. The FDE builds it.

Forward Deployed Engineer

The how - AI agent architecture, build, MCP integration, production deployment

Translates your requirements into working software. Flags what is technically feasible before you commit to a client.

What You'll Do

Process Discovery and Reengineering

Lead structured current-state process discovery sessions with client stakeholders - walkthroughs, observation, value stream mapping - at a fidelity that captures decision logic, exception handling, system touchpoints, and handoff points

Challenge the client's stated requirements with disciplined questioning: if a process step exists because 'we've always done it this way,' surface and resolve that before it becomes an automation constraint

Apply Lean thinking to identify waste, bottlenecks, and manual interventions that should be eliminated - not automated - before an agent is built

Produce redesigned future-state process maps that represent the optimized workflow the AI agent should execute, not the legacy process with AI layered on top

Facilitate executive-level process reviews to validate redesigned workflows and secure stakeholder alignment before build begins

AI Agent Requirements and Logic Design

Author clear, complete, and structured requirements for AI agent automations - written precisely enough for engineering and accessibly enough for business sign-off

Define agent decision logic in structured formats: decision trees, if-then-else flows, escalation triggers, exception routing rules, and confidence threshold criteria - without writing code

Document scope boundaries explicitly: what the agent handles autonomously, what requires human-in-the-loop review, and what must always be escalated

Write user stories and acceptance criteria that specify the actor, the action, the business outcome, and the measurable success condition

Identify and document edge cases, exceptions, and failure modes the agent must handle gracefully - derived from discovery, not assumption

Validate requirements completeness with the FDE before build begins - no ambiguity that would require the engineer to make business decisions mid-build

AI Platform Fluency

Develop and maintain working fluency in Claude Cowork and the Skills framework as a power user - understanding how Skills are structured, how agents reason through tasks, and how human-in-the-loop checkpoints are configured

Use Claude directly during discovery and validation - testing sample inputs against prototype agents to assess output quality before requirements are finalized

Distinguish between process steps best suited to AI agent automation, workflow tools, or continued human execution - and make that recommendation as part of the automation blueprint

Understand prompt design principles conceptually - sufficient to review and give feedback on agent logic without authoring the technical implementation

Recognize when a requirement is technically unfeasible for the current platform and escalate early rather than allowing the FDE to discover it mid-build

Stay current on Claude and broader AI platform updates - translating new capabilities into automation opportunities for clients

Business Case and ROI

Build automation business cases that quantify: current-state cost, projected AI-enabled capacity, error reduction value, compliance risk mitigation, and net investment payback period

Develop ROI models at conservative, base, and optimistic confidence levels - tied to measurable process assumptions the client validates

Prioritize automation opportunities by value-to-complexity ratio, enabling clients to sequence their AI roadmap by impact and feasibility

Produce executive-ready business case presentations that translate quantitative ROI into strategic narrative - what this automation enables the business to do that it could not before

Track realized value post-deployment against projected figures - closing the loop on delivery credibility

Stakeholder Management and Client Advisory

Design and facilitate discovery workshops, process reviews, requirements walkthroughs, and alignment sessions throughout the engagement lifecycle

Translate AI concepts - agents, Skills, prompt chains, human-in-the-loop design - into language that resonates with operations managers, compliance officers, and C-suite audiences

Manage competing stakeholder priorities and process ownership disputes that arise when automation redesigns cross departmental boundaries

Serve as the day-to-day business-side point of contact for client stakeholders, protecting the FDE's focus time for engineering delivery

Produce high-quality written deliverables: proc