Brian MacDonald

Senior Director of Product at Netomi building the agentic operating platform for CX at scale.

[Brian MacDonald @brianmac8500]

@RaminNasibov The Sweet Hereafter.

[Brian MacDonald @brianmac8500]

@DrNeilStone A bit like San Francisco. I've been there three times this year and honestly it's about the same as it's been since the 2000s. Same level of good and bad. Tenderloin even feels the same.

[Puneet Mehta @puneetmehtanyc]

This week at Cannes Lions, the world's leading CMOs are moving past the content creation conversation. The real discussion is the full customer journey. Stop building AI for each touchpoint separately. Build AI that reads the complete scene. What a customer browsed. What they bought. What frustrated them last time. What they need right now. When AI holds that full picture across every interaction, you stop reacting and start acting before they have to ask.

#CannesLions #CannesLions2026

[Brian MacDonald @brianmac8500]

@VancouverSun Just f**ING stop. All of you. Enjoy the bloody parades, the celebrations and the football. The world needs a little bit of joy right now.

[Puneet Mehta @puneetmehtanyc]

Deployment for AI, and making sure that you have risk management, sanctioning, and governance, are way more important than what raw models deliver.

@MetisStrategy #AgenticAI #CX

[Brian MacDonald @brianmac8500]

@parker_brydon We have approximately 40 people in Toronto. Perfect mix of talent and cost.

[Brian MacDonald @brianmac8500]

@netomi_official @Microsoft @Azure @MSFTDynamics365 We couldn't be more excited to be expanding our relationship with Microsoft. From Dynamics 365 to some exciting Data Fabric use cases, we see production-ready, agentic use cases emerging into the light.

[Puneet Mehta @puneetmehtanyc]

The model is rented. The architecture is owned.

That’s the line that will separate enterprise winners from losers in agentic AI - and it’s exactly what @satyanadella is naming. Enterprise orchestration, process context and the institutional learning loop have to live in the enterprise’s runtime, not the lab’s.

Architecturally, that means governance and context live inside the reasoning loop - context is what makes institutional knowledge compound inside the enterprise’s runtime; governance is what makes the model itself interchangeable.

We’ve been building this at Netomi and deployed across F500 like Delta, United, MetLife, DraftKings, Paramount. @OpenAI published us as the enterprise blueprint. @Accenture led our recent round because this architecture IS now the moat for enterprise AI.

[Puneet Mehta @puneetmehtanyc]

Good framing from @McKinsey and Parag Parekh at IKEA on where agentic CX is headed. The article suggests agentic AI for customer service isn't yet at phenomenal scale. My experience says otherwise. It is already in heavy use across the Fortune 500. The harder truth is that most deployments haven't scaled governance alongside transactions. Volume without policy enforcement at the moment of execution isn't scale. It's risk.

https://t.co/6AgYxIY7i8

[Puneet Mehta @puneetmehtanyc]

True for knowledge workers. Opposite is true for customers. A traveler whose flight just cancelled isn't explaining their intentions to an AI agent. The agent has to read the scene, apply loyalty rules, and deliver an allowable resolution in seconds. Intent, context, and policy simultaneously. Most enterprise architectures aren't built for that.

[Brian MacDonald @brianmac8500]

@noahkostesku It's a city of neighborhoods. The food scene is outstanding on all levels, be it Lake Inez, Bar Pompette, Dreyfus, Bernhardts. So much goodness in Markham and Scarborough. Dundas West, or Gehry, or even Leslieville can work. The parks, ravines, the Don, high park.

[Brian MacDonald @brianmac8500]

@evanlapointe I agree with you Evan. The "everyone is an IC" mantra is silly. Is every FDE going to be building a platform? A product? Probably some custom thing that gets used once. I love the acceleration but business, certainly B2B depends on finding the relationship and the fit

[Brian MacDonald @brianmac8500]

@levie Also, a great indication that whoever can productize the process and doesn't require a fleet of FDEs may also win.

[Brian MacDonald @brianmac8500]

@bradmenezes 100%. Exactly what we are seeing. It's all about expanding how the system of record, personalization, and memory can be used.

[Brian MacDonald @brianmac8500]

@signulll Live person

[Brian MacDonald @brianmac8500]

@20th_Centurygal Mogwai. Swans. Sunn o)))

[Brian MacDonald @brianmac8500]

@brettdash_ A fantastic alternative to NBA or prescriptive analytics. Does not replace the at-a-glance status that dashboards provide. I could argue for preemptive "business alarms," but it's easy to drown in these as well.

[Brian MacDonald @brianmac8500]

@sivareddyg @Mila_Quebec Incredible news!! Go Quebec!! (I've interviewed three PMs from Montreal recently, all exceptional)

[Brian MacDonald @brianmac8500]

@alliekmiller https://t.co/sPl5iFdAde

[Brian MacDonald @brianmac8500]

@adityaag I've been travelling to SF since the late 90s. Worked at Oracle for 12 years, and was in Redwood Shores every month or two. I had a gap between 2019 and Oct 2025 (start-ups, Covid) and been twice now. I fully agree with you, pretty much the same vibe with some bright spots.