Amber Gargrish

I started as the engineer. Now I lead the room.

I spent my first years as a full-stack software engineer at Infosys in India, building on .NET, Microsoft Dynamics ERP, and IBM BPM. That foundation still shapes how I lead: I know what it takes to build the thing, so I can plan it honestly and defend it credibly.

After moving to Canada, I grew into delivery leadership, then led a $15M+ AI/ML portfolio at Bell Canada. Today I lead agentic AI programs at Horizon 3 AI Labs, Unilever, where I have also originated and won four net-new programs this year.

I am also an Agile Coach with 6+ years of experience. The common thread is simple: make AI useful for real teams, govern it properly, and prove the value in numbers.

Eleven years, one direction: closer to the outcome.

Jan 2026 to present

Senior AI Program Manager and Delivery Lead

Horizon 3 AI Labs, Unilever, Toronto

Led an agentic AI ecosystem across a €3B global procurement portfolio with a €20M+ savings mandate: 10 concurrent projects, 8 product owners, executive reporting up to the CPO.
Saved 3,700+ man-hours through agentic automation and scaled adoption to 200+ users.
Originated and won 4 net-new AI implementation programs in 2026, pitching each and securing funding for the lab.
Transitioned the flagship program into operations and now leads a newer AI program.

4 of 4 pitches funded

Mar 2022 to Jan 2026

AI/ML Portfolio Lead

Bell Canada

Led a $15M+ AI/ML portfolio and a 100+ person cross-functional team.
Built and presented AI/ML business cases to the steering committee 15+ times.
Led enterprise incident management for S1 and S2 incidents: ran incident bridges, drove ownership and stakeholder cadence, and delivered process improvements.

Jul 2020 to Mar 2022

Delivery Manager

Corre Junction Inc.

Built costs, benefits, and KPIs for 50+ client business cases alongside project sponsors.
Led delivery of 30+ web, CMS, and e-commerce builds (WordPress, Shopify, WooCommerce) with budgets of $50K to $1M and timelines of 1 to 12 months.
Led a 10-person team to deliver a customer-facing learning management platform for training companies, and took a Canadian deals platform from kickoff to go-live in 6 months.
Cut projected delays by 20% through a delivery maturity assessment and improvement roadmap; ran monthly executive steering committees.

Jan 2019 to Nov 2019

Project Manager and Delivery Lead

Dogico Information Inc.

Delivered a data warehouse and Power BI executive dashboards for the CEO and VPs of Agnico Eagle Mines.
Authored 50+ PMO standards, including business case, SOW, SLA, and proof of concept templates, and led scope, budget, and estimation across 5+ project plans.
Led an 8-person cross-functional team to automate change management through a new service desk.
Governed schedule health against a 10% variance threshold, reporting to program governance and the monthly finance forum.

Jun 2014 to Dec 2017

Full-Stack Software Engineer

Infosys Ltd., India

Built and customized Microsoft Dynamics AX ERP modules (payables, receivables, general ledger, sales, inventory, warehouse, and trade and logistics) for Hitachi Powdered Metals, cutting invoice processing time by 25% and missed early-payment discounts by 20%.
Engineered warehouse layouts, picking strategies, and bin location logic, delivering 35% faster order fulfillment and 30% fewer picking errors.
Developed IBM BPM workflows for Wells Fargo across multiple lines of business.
Owned the path from requirements to UAT with business and QA teams: the foundation of how I scope, estimate, and defend delivery plans today.

.NETMicrosoft Dynamics AXIBM BPMSQLVisual Studio

Where I can help your organization.

Focused engagements for organizations in Canada and the US that want AI that is governed, funded, and actually in production.

AI governance and risk readiness

A governance model your regulators, auditors, and board can trust.

Governance frameworks aligned to OSFI E-23, NIST AI RMF, and ISO/IEC 42001
Model inventory, risk tiering, and accountability mapped to real owners
Readiness roadmap ahead of regulatory deadlines

Best for regulated organizations, especially financial institutions preparing for OSFI E-23. Discuss this

Proof of concept to production

Stop pilots from stalling. Give them a path to production.

A structured framework of documentation, checkpoints, and delivery gates
Implemented in 1 to 2 months, then owned internally as best practice
Clear go and no-go criteria leaders can act on

Best for teams with promising AI pilots and no repeatable path to scale. Discuss this

AI portfolio and delivery leadership

One view of every AI initiative, its cost, and its value.

Portfolio governance, prioritization, and intake
FinOps discipline for AI and cloud spend
Executive reporting that ties delivery to business value

Best for organizations running several AI initiatives at once. Discuss this

AI business cases and funding strategy

The case that gets your AI initiative approved.

Value framing, cost modeling, and delivery roadmap
Steering committee and executive pitch preparation
Built on a record of 4 of 4 recent pitches funded

Best for product and innovation leaders seeking investment. Discuss this

Workshops and speaking

Practical AI delivery and governance for leaders and teams.

The AI-ready project manager
Delivering agentic AI in the enterprise
AI governance for regulated industries

For conferences, universities, and internal leadership sessions. Invite me

Building in the open.

I am publishing my AI delivery and governance methodology as an open framework on GitHub, free for any team to use, adapt, and improve.

In progress, launching on GitHub

An open framework for taking AI from pilot to production

The same gated approach I use to move AI proofs of concept into governed, production-ready systems. Documentation templates, decision checkpoints, and governance controls mapped to OSFI E-23, NIST AI RMF, and ISO/IEC 42001.

Frame the valueProblem, owner, and measurable outcome

Assess the riskRisk tier and governance controls

Prove it worksEvidence-based proof of concept

Ready for productionOperations, monitoring, handover

Scale and sustainAdoption, value tracking, iteration

AI governance

What regulated AI governance looks like in practice

Turning frameworks into owners, controls, and evidence.

Agentic AI

Designing agentic workflows people actually adopt

Lessons from scaling to 200+ users.

Portfolio

Running an AI portfolio like an investment book

Prioritization, FinOps, and value tracking.

Off the clock, I sing.

Singing is my first love, and I accompany myself on guitar. Bollywood, Hindi, Urdu, and Punjabi music is where I reset, and it keeps me honest about practice: nothing good happens without reps.

I care about people, craft, and making technology feel useful rather than intimidating. Mentoring the next generation of delivery leaders is part of that.