Breaking Through Freight’s Invisible Friction

We recently sat down with Harish Abbott, CEO of Augment- an AI teammate for the logistics world that automates order-to-cash systems for freight brokers, fleets and shippers.

Harish’s background is rooted in technology applied to logistics. He began his career at Amazon in its early days, when the company only sold books. He later built several businesses, most recently called Deliverr, which created an Amazon Prime-like service for Shopify.

He shared the story behind Augment and why he chose to focus on this particular market.

Deliverr was a major buyer of freight, spending close to $75 million annually. Through that experience, Harish saw firsthand how tedious and error-prone freight procurement could be. As Deliverr grew and freight volume increased, the company had to scale its operations team just to keep up.

“For a single truckload between a broker, shipper and fleet, it's probably 15-20 emails, and potentially 4-5 calls that nobody knows about. That not only creates productivity doubts for people in this industry, but also delays the decision-making in your supply chain."

This fragmentation revealed a massive market opportunity—one that agentic AI could address across multiple pain points. As a consumer of freight, he realised how fragmented the process was.

His approach was to recreate the role of a human operations agent- only with perfect context of the organization on top of load level details- while freeing human workers to focus on higher-value, creative tasks.

“I actually looked at a job advertisement for an operations analyst and a freight broker- it was all about risk management, customer relationships- and then I’d look at the things they were doing and it was none of that.”

He believes recent advances in AI—particularly the ability to handle extremely large context windows—are enabling work that simply wasn’t possible a year ago.

 “Today, we are at a point where one of our customers has written us a sixty-page SOP for track and trace, and AI is able to handle that. And what that has done is that they now have a 40-person team that they could repurpose for customer success.”

Still, some capabilities remain out of reach—for now. An area, in Harish's opinion, where humans are very good at is unspoken judgement. There’s so much context that is tribal and so spread out in the company that it's almost impossible to document that today.

“I think what would be possible in the next six months is to extract this unspoken judgement from past decisions, emails, performances and try to get very close to that to at least assist a human.”

Overall, Harish is optimistic about AI’s future in logistics. The reality, he says, is that there’s an unlimited demand for intelligence in every corner. Humans can only analyze so much, but AI can help them analyze a lot more.

“I think we’re in a golden era of innovation. I’m so bullish that we will end up producing such amazing new sets of jobs that we don’t even imagine today.”

That said, job loss for some in the industry is inevitable. The key, he argues, is adaptation.

“Companies that are implementing AI, or personally even seeing AI around you, leaning in and learning that as a tool, preparing for your next set of  careers is the best thing you can do.”

In other words, while some roles will disappear, new ones will emerge. Who will be in the best position to get those new jobs? The people that learn AI as a toolset for themselves. 

Harish also noted that implementations fail not because of the technology itself, but because of weak integration with core systems. Without real-time context, AI cannot make reliable decisions — leading to frustration and loss of trust among users.

How should companies prepare to adopt AI successfully?

“You have to repeat the message, you have to train people, lean in and answer the sceptics, you have to find the champions. You have to do that hard work. It’s not the tech. I’ve seen enough successful implementations- the tech is there. It’s the integrations and the change management.”

His message to the industry is clear:

“The rate of change is not slowing down- it’s only going to accelerate, and we are in the very first innings of it.”

The rate of change is so massive that you can’t afford to wait a year. If you start today, you will have a head start and a leg up on this curve. 

What you can do is pick one simple use case to now have your people build the muscle memory on how to work with AI.

Just don't wait too long.

Listen to the episode on Spotify and Apple Podcasts, or watch the interview on YouTube.