Below is our recent interview with Tom Moore, CEO and Founder of ProvisionAI:
Interviewer: Could you provide our readers with a brief introduction to ProvisionAi?
Tom Moore: ProvisionAi ensures global companies’ supply plans are both high-service and can be cost-effectively executed…that is, making planning and execution “play nice together.” LevelLoad from ProvisionAi is a transportation scheduling solution featuring optimized, capacity-constrained replenishment. AutoO2 is an optimizing load builder that converts deployment requirements into efficient, damage-free shipments. Together, LevelLoad and AutoO2 drive customer service and long-term transportation planning objectives despite the realities of supply chain network constraints. The results include improved on-time and in-full customer service, fuller loads, higher use of preferred freight carriers, and lower costs. Clients save millions by tendering loads early to reserve preferred carriers, filling truckloads optimally to minimize wasted capacity, improving customer order fulfillment, and reducing costs. The patented technology is saving money and carbon for companies like Unilever, Baxter, P&G, and Kimberly-Clark.
Interviewer: Who is your ideal client and why?
Tom Moore: Should have a primary (Aka replenishment) freight spend shipping package products of >$2mm
Interviewer: Can you tell us something more about your solutions?
Tom Moore: Here are the problems we solve:
- Replenishment volatility. Spikes in shipping volume add significant cost in transportation and warehousing
- Because supply planning systems don’t consider the availability of space in shipping and receiving locations locations or the availability of labor replenishment plans can often be infeasible at worst, expensive at best.
- Poor OTIF (on-time, in-full)
- Replenishment loads are not full—wasting valuable payload capacity
The narrative looks like this:
Most supply planning systems tend to neglect the cost and network consequences of their plans. This oversight can lead to uneven shipment schedules from manufacturing plants to customer-facing warehouses, causing significant negative effects on operations. For instance, a sudden increase in trailer arrivals can result in insufficient space and labor to load or unload trucks, ultimately degrading customer fill rates as products needed for orders fail to be unloaded on time.
ProvisionAi’s innovative solutions harness the power of supply-planning systems and combine transportation data with the overall network’s constraints. LevelLoad performs daily optimization and generates a globally optimized replenishment transportation schedule, recommending early trucking capacity reservations across the entire network. To achieve this seamless flow, ProvisionAi employs standard operations research techniques and reinforcement learning—a cutting-edge AI approach—that effectively reduces transportation and warehouse costs while meeting customer fill-rate expectations.
At the same time, our AutoO2 optimizing load builder increases payload by 5-10%
LevelLoad from ProvisionAi is the only patented optimized replenishment transportation scheduling solution on the market. This next-generation transportation scheduling tool fixes replenishment effectively bridging the gap between supply planning and execution. This is a new genre of optimization that balances transportation cost and site capacities while meeting customer-service goals. The only competition is manual effort with Excel spreadsheets.
LevelLoad addresses the challenges of meeting the real-world constraints and cost pressures by analyzing shipment requirements set by the supply planning system for the next 30+ days and identifying hurtful spikes in demand, for example at month end. The system can then adjust by shipping some products early or holding less-needed items a day or two. This results in a more balanced transportation plan, enabling the use of preferred carriers and ensuring adequate storage space and labor availability across all sites.
Interviewer: What can we expect from ProvisionAi in next 12 months? What are your plans?
Tom Moore: In 2023, ProvisionAi removed 88,000 trucks from the road. Using AutoO2, companies can fill trucks fuller so that fewer trucks are used for replenishment. With 91% of all trucks underloaded, the time is now to max out trailers, keeping in mind that the loads need to be axle-legal and arrive without any product damage. Our clients increase payload by 5-10%. This year, in 2024, our goal is to remove 188,000 trucks from the road, helping to reduce carbon emissions.
Interviewer: What is the best thing about ProvisionAi that people might not know about?
Tom Moore: While the technology involves things like digital twins, operations research and AI, the real value is that it works. Companies such as Unilever and Kimberly-Clark can attest to:
- Savings – cost and environmental
- Increased carrier tender acceptance
- Higher OTIF
- Reduced carbon footprints for their and their carriers’ operations
- Lower carrier rates
- “Shipper of choice” status among carriers because of:
- Lower volatility
- Less deadhead miles
- Faster loading using AutoO2’s 3D load diagram
The patented technology:
ProvisionAi developed a compound approach that could take on the deployment smoothing challenge on each lane and simultaneously smooth the whole network and every location. A classic mathematical programming (Operations Research) technique could balance flows across days and respect both carrier and site limitations, but could not efficiently answer part of the question: how much high-priority product would fit on each potential truckload that we might decide to tender to carriers? The math program needed a way of building good loads very quickly to enable a holistic solution.
Enter AI. Using Reinforcement Learning, LevelLoad swiftly constructs realistic candidate loads and know that bricks can’t be stacked on eggs and to “max out” the loads it is important to understand the weight/cube mixes of products. Working iteratively between AI and mathematical programming, the ProvisionAi LevelLoad system builds an optimal deployment plan for the next 30+ days – smoothed on every lane, staying within site capacity limits, and ensuring the highest deployment-priority products are shipped – no customer service failures.
LevelLoad employs a cutting-edge combination of linear programming, heuristics, and reinforcement learning (AI) to achieve these results. The system processes network demand, order, forecast, and transportation data, along with facility constraints, to generate daily optimizations. The output is a globally feasible, cost-effective, and customer-service satisfying replenishment transportation schedule. This enables shippers to commit to carriers earlier and build optimized shipments, streamlining the entire supply chain process while delivering high customer service.
3D load design in AutoO2 enables sites to ship planned items for each truck, decreasing transit damage, reducing landfill waste, and improving on-shelf presence.