“As AI scales in supply chains, human decision-making shifts from controlling flows to architecting systems—focusing less on what to do and more on when to intervene, what risks to take, and how to balance trade-offs at scale,” highlights Akhil Srivastava, Senior Director – New Business Development, International Supply Chain & Innovation, AB InBev.
While AI adoption in supply chains is still evolving in markets like India, how do you see the role of human decision-making changing as these technologies scale?
As we see AI adoption scaling across supply chains—especially within vibrant emerging markets like India—it’s clear that the role of human decision-making isn’t diminishing; it’s evolving and moving upward, transitioning from transactional control to strategic orchestration. The narrative isn’t about replacement, but rather a redefinition of where humans truly add value.
Let’s look at how this shift unfolds. Today, in environments with low AI maturity, humans are at the center of planning, forecasting, and inventory decisions, relying heavily on experience, spreadsheets, and heuristics. As AI becomes more integrated, it takes over demand forecasting, replenishment triggers, and route optimization, freeing people to focus on managing exceptions, validating model outputs in volatile contexts, and ultimately, making strategic trade-off decisions. In markets like India, where volatility and infrastructure constraints are real, this shift is particularly pronounced.
AI also revolutionizes data processing. Where leaders once asked, “What is happening?” or “Can I trust this data?”, they now move toward interpreting what the data means for network strategy, identifying risk signals, and contextualizing AI outputs with local truths—like festivals, regional demand spikes, and policy shifts. Humans become indispensable as context interpreters and reality validators.
While AI optimizes individual nodes—such as warehouse efficiency and transport costs—human decision-making increasingly focuses on cross-functional trade-offs: inventory versus cash flow, service levels versus cost-to-serve, network-wide optimization, and long-term capability building such as localization and supplier diversification. In this landscape, humans act as system architects, not merely node managers.
As AI brings predictive models, simulations, and digital twins into the mix, planning transitions from deterministic to probabilistic and scenario-based. Human roles shift towards interpreting probability distributions, making decisions amidst uncertainty, and navigating ambiguity—areas where local volatility and supplier fragmentation in India make human expertise crucial.
Another critical evolution is from control to governance and ethics. As AI scales, humans must define when it can auto-execute, when it must escalate, and set acceptable thresholds for stockouts, forecast errors, and cost deviations. Leaders become guardians of decision frameworks, ensuring AI remains unbiased and avoiding “automation complacency.”
As AI adoption accelerates, decision-makers need to build new capabilities across four dimensions: data and model literacy (understanding what models do and their limitations), scenario thinking (embracing a “what if” mindset for disruptions), commercial acumen (translating AI recommendations into P&L impact), and change leadership (driving adoption in teams and bridging frontline with digital systems).
Ultimately, AI is set to industrialize decision-making at scale—but humans will define the rules, intervene strategically, and carry higher accountability for each decision. Fewer manual tasks, more strategic interventions, and an elevated role for human judgment are on the horizon.
Which traditional supply chain roles are losing relevance, and what new roles or capabilities are emerging as critical in this shift?
Traditional supply chain roles are evolving in response to technology-led changes. We’re not seeing jobs disappear entirely, but rather, many traditional roles are shrinking, being automated, or transforming into more complex and valuable work.
For example, transactional planning roles like demand planners who rely on spreadsheets, or supply planners running manual replenishments, are becoming less relevant. AI-driven forecasting and autonomous planning systems are outpacing manual, rule-based approaches, and continuous planning is replacing periodic cycles. The nature of these roles is shifting from plan creation to plan orchestration and exception management.
Order management and execution coordinator positions, which require manual order entry, allocation, and shipment tracking, are also declining. With RPA, EDI/API integration, and control towers automating workflows and execution visibility, and real-time logistics data reducing the need for human tracking, these roles are shifting from execution follow-up to network control and disruption response.
In pure procurement transaction roles, tactical buyers who focus on RFQs and PO processing are seeing their work streamlined by e-sourcing platforms and AI-based vendor comparison tools. Automated catalog buying and replenishment are reducing the need for manual involvement, with the focus moving from buying to supplier strategy and risk management.
Static reporting and MIS roles, where analysts generate historical reports, are also affected. Self-serve BI tools and AI copilots now deliver instant insights, so the role is evolving from reporting to decision intelligence and insight shaping.
Even narrow functional specialists—those who are siloed in logistics, warehousing, or planning—are being affected. The demand now is for cross-functional thinking, as end-to-end visibility and integrated planning become more critical. The shift is from pure functional depth to systems-level thinking.
The real transformation is in the mix of capabilities. The most significant change isn’t necessarily in job titles, but in the skills required. Where we once focused on forecast accuracy and process adherence, scenario agility and decision-making under uncertainty are now essential. Functional expertise is giving way to end-to-end systems thinking, and cost optimization is being replaced by a focus on risk and resilience optimization. Manual analysis is evolving into AI-augmented decision making, and reporting is transforming into insight storytelling.
It’s important to note that around 40–60% of traditional transactional supply chain work is automatable. The value is migrating from execution to orchestration, efficiency to resilience, and data generation to decision intelligence.
Organizations that succeed in this new landscape will flatten transactional layers, build decision-centric roles, invest in AI-human hybrid capabilities, and redesign their structures around process flows instead of traditional functions.
Let’s keep these shifts in mind as we continue to build our teams and develop our skills for the future.
What are the most essential skills the next-generation supply chain professional must develop to remain relevant in an algorithm-driven environment?
As we continue to navigate the evolving landscape of supply chain management, I wanted to share some key leadership skills that are emerging as critical in our increasingly digitized environment. These skills not only enhance our ability to leverage technology but also strengthen our capacity to drive business outcomes and adapt to change.
First and foremost, data fluency and analytical thinking have become non-negotiable. Being able to interpret, challenge, and act on AI and algorithm outputs is essential. A strong grounding in statistics, forecasting logic, and optimization principles, along with comfort using tools like Python, SQL, or BI dashboards, equips us to validate context, bias, and trade-offs in the insights generated by algorithms. Moving from simply consuming reports to owning decisions, and asking questions such as, “What assumptions is this model making?” helps us stay proactive.
Next, systems thinking and end-to-end visibility are vital. Understanding supply chain as an interconnected system—from demand and supply through logistics to the customer—allows us to anticipate ripple effects and optimize system-wide performance. Identifying trade-offs among inventory, service, cost, and cash is key to preventing suboptimal “black-box optimization.”
Human and machine decision orchestration is another important area. Knowing when to trust the algorithm and when to override it and designing frameworks that combine AI recommendations with human judgment, ensure we strike the right balance. Escalation logic for exceptions and clarity on decision rights (what’s automated and what’s manual) are practical tools that help manage this balance.
Digital and AI literacy is essential, but it needs to be applied rather than theoretical. Understanding the capabilities and limits of AI/ML forecasting, digital twins, control towers, and autonomous planning systems positions us to deploy and scale technology effectively. Acting as a bridge between operations and data science teams and translating business problems into technical solutions drives real impact.
Scenario planning and risk thinking develop our ability to model uncertainty and disruption. Strong grounding in probabilistic thinking helps us build contingency playbooks and stress-test supply chain resilience, recognizing that algorithms are often trained on historical data and may not always respond well to non-linear or novel disruptions.
Execution excellence in a digitized environment means driving outcomes in automated, real-time supply chains and managing exception-based operations. Value increasingly shifts from simply doing tasks to managing overall system performance, so focusing on cycle times, service levels, and cash velocity—and operating via dashboards and triggers—ensures we keep pace.
Cross-functional influence and stakeholder alignment are also important. Aligning commercial, finance, manufacturing, and logistics stakeholders, and translating analytics into business language, helps drive adoption of algorithmic decisions, recognizing that leaders—not algorithms—manage stakeholders.
Change leadership and adoption management are crucial for transformation. Leading the shift from legacy processes to digital workflows, managing resistance to automation and AI, and training teams on new decision models build trust in algorithm outputs and ensure that technology ROI is fully realized.
Commercial and financial acumen strengthen our ability to understand the P&L impact of supply chain decisions, linking operations to growth, margins, and working capital. This positions supply chain as a strategic profit lever, not just a cost center, and enables us to translate service improvements into revenue impact and optimize cost-to-serve.
Finally, agility, a learning mindset, and curiosity are indispensable in a rapidly evolving tech environment. Continuous upskilling, comfort with ambiguity, and rapid iteration ensure we remain competitive, as tools change faster than roles and our learning speed becomes the ultimate differentiator.
How are organisational structures and leadership models evolving as supply chains move from process execution to system orchestration?
Core issue: The gap isn’t just “AI skills”—it spans three distinct layers:
Build a “T-Shaped” Workforce (Most Effective Mode) with Broad AI/Data literacy across all supply chain roles and Deep expertise in a few specialized teams
- Create “Use-Case Driven Upskilling”
- Instead of generic AI programs, anchor learning to high-impact supply chain use cases:
Looking ahead, what will differentiate supply chains in an AI-driven future, and how will talent capability shape and accelerate this growth?
In high stakes situations (safety, legal exposure, large financial impact, reputational risk), the right balance is not “AI vs. human” but system design: create conditions where AI accelerates reasoning while humans retain accountable authority, supported by guardrails that make over trust and under trust both unlikely.
Make AI “trustworthy by design” using calibration, not persuasion. Humans are prone to automation bias (over-trusting AI) and algorithm aversion (rejecting AI after a visible error). Calibration reduces both: people trust AI appropriately when it “knows what it doesn’t know.”
Our approach to AI oversight ensures our systems continue to deliver value while upholding high standards of safety and ethics. First, it's crucial that human judgement is focused on the right areas, namely context, ethics, and exception handling. By concentrating on human expertise where it's most impactful, we can trust AI with pattern detection, scenario simulation, early warning signals, and optimization within known constraints.
When it comes to monitoring performance, we should rely on leading indicators, not just outcomes. Waiting for a failure means we've already missed an opportunity to intervene sooner. Instead, let's proactively track signs such as calibration drift (where confidence and actual accuracy diverge), data drift (shifts in input distributions), override patterns (whether humans are overriding AI decisions too often or too rarely), near misses (instances where guardrails successfully averted negative outcomes), and decision latency (observing whether AI is speeding up decisions without sacrificing quality).
Ultimately, a mature organization treats AI as a “living system”—something that requires ongoing attention and adaptation, rather than a static feature that's simply deployed and forgotten. By adopting these practices, we can ensure our AI capabilities remain robust, ethical, and aligned with our organizational goals.
In an AI driven future, supply chains won’t be differentiated by assets or even scale—they will be differentiated by decision quality at speed, enabled by intelligent systems + augmented talent. The winners will be those that combine machine precision with human judgement at the right control points.
Below is a clean, executive view of what will differentiate supply chains and how talent capability becomes the multiplier.
- From linear execution - Autonomous, self-orchestrating networks.
- Not visibility—but closed-loop execution (detect - decide - act - learn). Ability to price risk into every operational decision in real time.
- Predictive + pre-emptive resilience, not reactive firefighting when supply chain becomes a margin engine, not just a cost center
In summary, humans move from doing tasks to designing and supervising decisions.
Categories
