Webinar Recap: Eliminating Waste in Supply Chains — Unlocking Efficiency with AI-Powered Decision Intelligence

Summary

In our Future.Now webinar, “Eliminating Waste in Supply Chains: Unlocking Efficiency with AI-Powered Decision Intelligence,” we explored how enterprises can curb excess and obsolete (E&O) inventory while boosting performance. The session framed why inventory waste persists — rising complexity, volatile demand, and fragmented processes — then showed how decision intelligence changes the game by detecting surplus risks early, evaluating trade-offs, and recommending targeted actions before waste escalates. We also touched on governance and adoption — how teams vet recommendations, capture rationale, and measure effect — so improvements stick beyond a one-off cleanup.

We discussed industry patterns (excess vs. obsolescence), the capital and service impact of overstocking, and practical examples where decision intelligence unlocked value in weeks. The live demo walked through how Aera, the decision intelligence agent, continuously monitors material, production, and inventory signals; simulates options against constraints; and can execute actions to throttle supply, shape demand, reallocate stock, or adjust planned orders. We closed on outcomes of implementing the agent: reduced write-off risk, lower working capital, and measurable sustainability gains through source-level waste reduction — at scale and with consistent control.

Key Takeaways

• Inventory waste is structural — and solvable.
  Disruptions, rapid product cycles, and network complexity raise the odds of overstock and obsolescence. Decision intelligence addresses this by continuously scanning live signals — demand, supply, lead times, and policy thresholds — surfacing surplus risks early, and guiding precise responses that prevent cost from building up. Crucially, it standardizes how organizations spot, prioritize, and resolve risks across brands and regions.
• Excess and obsolescence require different plays.
  Excess often stems from over-ordering, location imbalances, or slowing velocity; obsolescence follows design changes, regulatory shifts, or quality issues. The right response varies — reallocation to under-served markets, demand shaping with targeted promotions, or plan adjustments and buy-downs — so you need a system that understands context (SKU, customer, constraints) and proposes the right move with quantified impact and confidence.
• Targeted actions beat blanket cuts.
  The examples showed how repurposing or reallocating inventory, rather than broad reductions, frees capital and avoids write-offs while protecting customer commitments. Decision intelligence recommends specific moves (what to change, where, when), simulates alternatives, and shows expected outcomes on service and margin. Teams can compare options side-by-side, approve the best one, and track realized benefits for continuous improvement.
• Live data + simulation improves outcomes.
  Aera evaluates choices before recommending action, accounting for capacity limits, material availability, customer priorities, and planning horizons. This reduces firefighting and second-guessing while improving service. Over time, the agent learns from decision outcomes, refines thresholds and heuristics, and raises recommendation quality — turning tribal knowledge into repeatable, governed practice.
• From recommendations to action — at scale.
  Beyond analytics, the agent can execute approved actions and write back to source systems, shortening the loop between insight and result. This creates consistent governance, auditable decisions, and scalable improvements across SKUs, sites, and regions without adding manual workload. Organizations move from periodic cleanup projects to a continuous, closed-loop process that sustains results.

Speakers

Jeroen Nysen, Director, Supply Chain Analytics & Intelligence, Deloitte. Jeroen brings close to 12 years of experience in aligning supply chain needs to the right digital capabilities. He leads Deloitte’s supply chain analytics and intelligence offering, specializing in control tower, decision intelligence, and guiding multinational clients across industries to make faster and better decisions. He has collaborated with Aera Technology for 6 years, coordinating the Deloitte–Aera alliance for the EMEA market.

Archana Ravi, Director, Growth and Solution Engineering, Aera Technology. Archana brings 17 years of experience in supply chain strategy, operation, and product management experience across industries and consulting. She has partnered with some of the world's leading companies in food and beverage, retail, CPG chemicals, and pharma leaders to transform and optimize their complex supply chain. She works closely with Aera Technology customers to help them realize the full value of decision intelligence, guiding them to define their North Star vision, accelerate their digital transformation journey, and turn big aspirations into real, measurable results.

Full Recording

Access the full webinar recording here.

Q&A

Q: The presentation showed impressive results. How long did it take to achieve these results, and what were the key success factors?

Jeroen Nysen: It depends on the use case, but with decision intelligence on the Aera platform the time to value is really impressive — especially compared with more traditional, large APS or transactional systems that follow a long, waterfall approach over several years. Here it’s very dynamic: we’re talking several weeks to reach a working product, which can then be further enhanced with additional capabilities. The time to value here can be measured in weeks, not months or years.

Archana Ravi: Adding to that: beyond quick time to value, the pre-built Decision Data Model (DDM) accelerates ingestion and decision-logic validation. The platform’s composability (its modular design) lets you build skills tailored to your business without starting from scratch. These are key success factors that enable teams to get going quickly on the platform.

Q: The presentation emphasizes that “planning alone is not the solution.” How does decision intelligence complement traditional supply chain planning tools?

Archana Ravi: Aera is deployed for many customers who already have a planning tool in place, and the two go hand in hand. First, planning typically operates in time horizons beyond the frozen period. You deploy decision intelligence within the frozen horizon, where a lot of changes are happening and you need to respond quickly. Secondly, you need to be able to pull data from different source systems. For example, one customer of ours used planning, order management, distribution, and warehouse management data together to make the right decision. Working across systems to make the right decision is a key differentiator here.

Jeroen Nysen: We see decision intelligence complementing planning (such as handling deviations in the frozen period with exception-based management to course-correct) and also augmenting planning altogether. For example, dynamically setting planning parameters such as lead time: some APS solutions bake in fixed lead times, but lead time fluctuates over time and across scenarios. Without reflecting those fluctuations, you get inappropriate planning and multiple disruptions. So you can use decision intelligence both to deal with disruptions and to enhance planning so fewer disruptions are generated. In the end, they’re truly complementary sets of capabilities.

Q: How do you determine which decisions should be automated vs. augmented vs. supported? What framework do you use?

Archana Ravi: As Jeroen briefly touched on, the product supports both structured decisions and situational decisions. For high-frequency, highly-granular decisions that pull from different sources, you typically deploy a skill or a decision agent for automation. Aera also supports situational decisions through its calc capabilities, which enable modeling when there are tactical or strategic decisions that you need to take, such as in long-term capacity planning.

We didn’t have time to dive into the other capabilities today, but we’re happy to connect offline and walk through them. In practice, you automate frequent, data-rich decisions with skills and augment/support tactical or strategic decisions with modeling and human review.

Q: What’s the biggest mindset change teams need to make for this kind of transformation?

Jeroen Nysen: A couple stand out. First, it’s important to treat decision intelligence as a principle and a capability, not a single use case. We highlighted several use cases — inventory aging, allocation, health management, and others — but the real opportunity is to think about decisions across the end-to-end value chain and orchestrate them so you automate and augment decision-making to reach better, faster, and more decisions overall. Don’t get stuck on one use case.

Second, it’s important to rethink readiness. Many organizations feel they must first achieve full data availability or end-to-end visibility before they’re “ready.” That can become an endless cycle. Instead, work backwards from the decisions you want to make — where value is leaking or opportunity is largest — and identify the data and other requirements needed to unlock that decision-making process. You’ll achieve faster time to value than with a sequential, waterfall approach that waits for each prerequisite. Keep a value-based mindset, focus on the desired outcome, and build backwards from there.

Q: How does data from our ERP system integrate into Aera?

Archana Ravi: Data is transferred securely through Aera’s patented agent architecture. Predefined crawlers extract data from ERP systems and can ingest thousands of data fields and tables, including: master data (e.g., materials, vendors, customers), Transactions (e.g., orders, shipments, inventory), financials (e.g., billing, cost centers), and procurement and planning data. Aera can also ingest unstructured data via the agent architecture. The agent also supports bidirectional communication; it can read data from and write approved decisions back into your ERP. Once ingested, ERP data is mapped and cleansed using Aera’s step-based stream flows and harmonized into a single, unified Decision Data Model, enabling cross-functional visibility and real-time context.

Q: What challenges have you faced in customers adopting AI and embedding it into their day-to-day workflows? How much time do you typically need to support customers before they can use the system independently?

Archana Ravi: Customers typically face challenges with master data quality, trust and change management, and system integration. Aera’s platform and methodology address these from day one:

• Master data quality: Many organizations begin with inconsistent or inaccurate master data across ERPs and source systems. Aera’s Decision Data Model and Master Data Skill continuously monitor, cleanse, and harmonize data using embedded business rules, creating a trusted foundation for decision automation.
• Trust, change management, and adoption: Aera’s “glass box” approach builds user trust through full transparency — each recommendation includes contextual information, expected business impact, and a confidence score.
• Integration simplicity: Aera’s crawler technology connects seamlessly with ERPs, planning systems, and data lakes, providing real-time, bi-directional data flow. This enables faster onboarding, minimal IT dependency, and continuous synchronization across systems.

Together, these capabilities help customers quickly build confidence in Aera’s decisions, reduce manual effort, and achieve scalable adoption typically within four to five months.

Jeroen Nysen: To all that I would advise…

• Ensure fit-for-purpose through early involvement: Make sure designed and implemented decision-intelligence use cases are tailored to actual end users and decision-makers so they can make better decisions. Involve them from the very beginning in ideation and design, not only later at UAT.
• Ensure buy-in through a value-oriented mindset: Keep value realization up front in ideation and design, and clearly measure and display realized benefits after go-live.
• Ensure trust through logic transparency: In reference to Archana’s glass box principles — make what the AI does easy to understand, and able to be tested and reviewed.
• Ensure motivation through a seamless experience: In reference to integration simplicity and overall ease of use — aim for the least number of clicks from issue detection to resolution.

Q: Do you offer supply chain accelerators to integrate with the main supply chain processes or areas?

Archana Ravi: Yes. Aera offers a robust set of prebuilt supply chain accelerators designed to integrate with and enhance core supply chain processes — across planning, inventory, and fulfillment. These accelerators help organizations achieve faster time-to-value while maintaining flexibility to adapt to specific business needs. Aera Decision Cloud is built as a modular platform where decision logic, data, AI models, and user engagement are separate but interoperable components. You can reuse and adapt prebuilt skills across use cases, functions, and geographies, and rapidly compose new skills by combining existing decision components and leveraging platform capabilities.

Jeroen Nysen: Just as in Aera, Deloitte has built a repository of prebuilt supply chain accelerators and skill templates. These help organizations get started from an initial skill architecture, which can later be fine-tuned and tailored to a client’s specific needs. Deloitte also brings deep supply chain and industry expertise to adapt these templates and standards accordingly.