Jacky Zhao 🚚🤖 on LinkedIn: A great example of using AI to collectively improve the industry… (2024)

🚀 Suuuuuuper excited to show off our "Appointment Scheduling" autonomous agent! Not all AI agents are built equal. Our enhanced agent is fortified to handle issues effortlessly and provide a seamless scheduling experience. Take a peek at our demo below: The truck driver ran out of hours and needed to push his delivery time back. The AI agent coordinated with the drop-off facility and the driver to get the shipment back on track![sound on 🔊]#freight #logistics #transportation #ai #freighttech #freightbroker #trucking

25

2 Comments

📢 #freightbrokers 🚨 I need your help, please!The FreightStack AI team is stress-testing our newest multi-party 'Appointment Scheduling' AI agent.Example: A broker coordinating pick-up and drop-off times with a driver and multiple facilities.Could you share the most challenging scenarios you've encountered while scheduling appointments?#freight #freighttech #freightbroker #carriers #dispatcher #logistics #transportation

11

14 Comments

I'm stoked about Truckstop's new tool, which seamlessly integrates carrier compliance and risk factors directly into the side panel. Had a blast partnering with the Truckstop crew to make this happen. FreightStack AI is committed to fighting freight fraud and we're proud to power this initiative with the FreightStack AI engine.

36

3 Comments

Anthony Chen and I had an INCREDIBLE time sharing our story with Nate Shutes 🙌 If you’re passionate about building solutions in the logistics space, be sure to subscribe to Bootstrapper's Guide to Logistics for more founder insights and stories!

37

2 Comments

I've been thinking a lot about what the future of the workforce looks like in five years (maybe even sooner).🔵 Current State of AI Agents: Currently, AI agents function more like "interns"—very booksmart but with limited reasoning and decision-making capabilities, and without any real-world experience.>>> They follow instructions without questioning the rationale behind decisions.🔵 Future State: As models improve, they will begin to operate more like "senior-level employees". These advanced AI agents will not just execute tasks but also provide valuable feedback and suggest better alternatives based on past experiences and best practices.>>> Instead of simply agreeing to a proposed marketing strategy, a better AI might suggest a more effective approach based on recent data trends.🔵 Orchestrators of Work: Employees will evolve into "orchestrators of work" rather than "servants of work" to optimize workflows. This shift empowers employees to take on more impactful roles within their organizations.>>> An employee might oversee multiple AI agents coordinating a complex project, ensuring strategic alignment and creative problem-solving.

20

3 Comments

Re: Copilots vs Autonomous AgentsI believe AI automation is a sliding scale. Anyone who has worked a day in logistics will tell you that you need a human in the loop for most workflows. Simple, tedious tasks can be largely automated (like automatic quotes, track/trace, etc), but the sheer complexity of freight requires a human to be in the driver's seat.The simple reason is this...AI agents (in their current state) are like extremely smart interns. They are very book smart, able to retain and spit out knowledge with ease, but generally green around the nuances of the business. This WILL improve over time as newer models develop better reasoning capabilities.I prefer the copilot approach as it combines the strengths of AI and human oversight. The AI handles repetitive tasks, data processing, and offers suggestions based on patterns, while the broker oversees operations, makes critical decisions, and handles complex exceptions. This means faster processing, fewer errors in routine tasks, and allows brokers to focus on strategic activities and problem-solving.Copilots empower freight brokers to become more like "orchestrators of work".

16

1 Comment

It's SUCH a fascinating time to be building solutions! I think the last time engineers were this stoked with excitement was during the rise of mobile apps.The reason is that LLMs (Large Language Models) have unlocked three core building blocks (primitives) that were previously nearly impossible to achieve consistently and at scale:1) Reasoning and Decision Making: The capability to choose water over bread if you are thirsty.2) Semantic Search: The ability to understand the intent and context of a query.3) Data Transformation: The ability to convert input data from one form to another (e.g., unstructured to structured, books to summaries, JSON to SQL).These primitives provide new ways to tackle problems that were once thought to be insurmountable. The future is wide open with possibilities, and we are better equipped than ever to build it.

17

3 Comments

If you're building, you should conduct an intensity and frequency analysis. Anthony Chen and I wanted FreightStack AI to address problems in the "high intensity" + "high frequency" quadrant.HIGH FREQUENCY:Most broker reps wake up to over 300 emails a day, each with competing priorities. David Bozeman (C.H. Robinson) mentioned that they receive 50k emails a day, while Doug Waggoner (Echo Global Logistics) said they receive 60k emails a day. Brokers need to read each email from a customer/carrier, research the question across multiple systems, and then respond with an answer. Too many workflows, varying priorities, and recency bias means that brokers are spread too thin and have no time to build their book of business.HIGH INTENSITY:Over 30% of emails are not responded to within the first hour, where timing is everything in this business. The majority of accounts are managed only by exception, thus having a higher risk of churn. Humans are fantastic at handling nuanced complexity and managing a few tasks well, but not at multi-threaded parallel work streams. This results in only the top 10% of accounts receiving focused attention. This issue is compounded by the fact that it takes an average of 4-6 months to train a new broker, yet the average retention rate for a class of broker reps is approximately 20% by the end of the year.

34

1 Comment