In the heart of Chicago, at the Journal of Commerce Inland Distribution Conference 2025, industry discussions centered intensely on how technology is reshaping freight operations. One of the most crucial topics was the strategic implementation of Artificial Intelligence (AI) in freight brokerage, moving AI from a niche tool to an operational necessity.
Annalise Sandhu, Co-founder and CEO of Chain, an AI assistant for Freight Ops TMS, provided crucial insights into how brokerages can navigate this rapidly evolving landscape. Chain focuses specifically on automating manual work in the track and trace process, including check calls, ETA updates, exception detection, and document collection.
As brokers explore adding automation, two primary questions consistently arise: How do I evaluate different AI providers, and where is the Return on Investment (ROI)?.
Regarding provider evaluation, Annalise noted the overwhelming amount of "noise" in the AI space today. Her advice is straightforward: utilize customer references. Brokers should connect with others in their network—especially those with similar business models—to understand how they are applying AI to their business and what results they are seeing.
The second major question, concerning ROI and the future of current employees, is not as simple as some might think. Annalise stated, "I think the best way to do that is just with customer references.". The expectation that applying AI will result in the immediate layoff of an entire team is often unrealistic. Instead, the change is often seen among faster-growing freight brokerages that can expand their teams and take on more customers without needing to proportionally hire and scale their track and trace personnel.
So, what is AI uniquely qualified to handle within a brokerage? According to Annalise, AI performs exceptionally well at the most mundane and redundant tasks. While simple automation handles some tasks, AI’s strength lies in dealing with messy, unstructured data and conversations.
Annalise explained that AI eliminates the need for employees to constantly perform repetitive outreach, stating, "A human shouldn't have to reach out... and ask the same question over and over like okay are you going to be there on time is your ETA still good...".
Logistics remains disconnected, with various systems in use. If people provide updates like ETAs, arrivals, or departures, AI can easily consume that data and update the TMS automatically. This capability has become increasingly viable because industry systems have become much more open, implementing and taking API accessibility seriously. This ensures the AI has access to the same crucial data—from the TMS to tracking providers—that team members use to make proper decisions.
Historically, one of the earliest areas of successful automation traction in the brokerage space was automated pricing, which focused on increasing revenue. Today, the focus has shifted, as labor costs are rising and freight rates are low due to the persistent freight recession, squeezing margins from both the top and bottom.
Brokerages are now looking to maximize their existing margins by addressing operational costs. An industry expert noted this trend, observing that many AI tools now focus "not necessarily winning you new revenue, but like reducing your cost to operate so that the revenue you have is maximized.".
For companies like Chain, this means helping brokerages scale efficiently. One customer, for instance, saw their revenue per headcount jump from $2 million to $4 million after implementing Chain, achieving significant growth with the same headcount. The implementation is a strategic, long-term investment, moving beyond simple, immediate ROI calculations. As one industry expert commented, regarding the current necessity of automation, "you're almost crazy if you're not looking at it.".
As 2025 wraps up and brokerages finalize their 2026 priorities, Annalise offered critical advice for those exploring automation solutions.
First, maintain an open mind toward exploring different solutions for educational purposes. Second, and most importantly, understand the specific use case.
Annalise warned against implementing AI simply because everyone else is doing it. Brokers, especially those in the C-suite, must sit on the floor, observe employees, and identify precisely where team members spend the most time on repetitive tasks.
The key to a successful AI strategy is pinpointing where automation will create the most impact on the business. For some, this might be check calls; for others, it might be different pressing issues. Understanding the employees’ workflow is essential before committing to a provider.