Latest news with #ShekarNatarajan
Forbes
2 days ago
- Business
- Forbes
The Age Of One: How AI Can Personalize The Supply Chain For Every Individual
Shekar Natarajan is the founder and CEO of Examine your daily life, and chances are, you'll see many cases of personalization. Launch Netflix on your TV, for instance, and you'll get tailored content recommendations. Open the Uber Eats app to order something to eat while you watch one of the recommended TV shows or movies, and a selection of restaurants the algorithm thinks you might enjoy will show up on your screen. These are just two examples of many that demonstrate how personalization has become commonplace. In fact, research published in 2021 by McKinsey reveals that consumers want personalization. Specifically, 72% of surveyed consumers 'said they expect the businesses they buy from to recognize them as individuals and know their interests.' McKinsey's research, in my view, indicates what I believe the future holds—an increased need for companies to embrace the business-to-me (B2Me) approach rather than the business-to-consumer (B2C) one. However, the current supply chain industry is not well-equipped to usher in the dawn of B2Me. That's where AI can step in. The Limitations Of Current Supply Chains When It Comes To The B2Me Approach The B2C approach works in some cases. Certain companies, namely those that produce mass market products such as toothpaste and soda, target aggregate customer segments. Personalization arguably isn't necessary for mass market products. On the flip side, the B2Me approach is rooted in personalization and helps facilitate consumer loyalty and retention, as opposed to more transactional relationships. For example, the ability to order personalized clothing and backpacks from a company based on your style preferences involves you more deeply with that company—you're getting products that are specifically designed for you. The state of today's supply chain industry is ill-equipped for the B2C approach, let alone the B2Me approach. Supply chains were, by and large, built for the business-to-business (B2B) world—efficiency, cost reduction, stability and inventory planning—not for creating a great customer experience and offering value-added functions. From my observations, these characteristics often struggle to translate to the B2C approach in practice, but in theory, they should be effective. When you're mass producing products such as soda and toothpaste, rather than personalizing them, you want a supply chain environment that's set up for efficiency, cost reduction, stability and inventory planning. But under a B2Me approach, these characteristics are weaknesses because they don't prioritize the customer experience and value-added functions. Additionally, many supply chains were designed with outdated systems intended for planning, rather than real-time responsiveness or proactive action. These outdated systems often fail to integrate well with other tools, such as CRM solutions, resulting in fragmented data and disorganized communication. Under a B2C approach, these are limitations, but not necessarily consequential limitations, because products are mass-produced. Under a B2Me approach, these limitations are quite severe. AI Can Help Create Empathy In Supply Chains That Power The B2Me Approach AI can help build empathetic supply chains that power the B2Me approach. It might sound strange to associate supply chains with empathy, but what I'm referring to are supply chains that understand people's needs and contexts and can adapt in real time. Supply chain leaders shouldn't think of AI as an automation capability. Instead, they should view it as an orchestration capability. AI can enable supply chain leaders to integrate the various components of their logistics systems, allowing them to offer more personalized products and/or services. I mention both products and services because, in my view, personalization doesn't necessarily have to refer to products (some products, such as toothpaste and soda, don't necessarily make sense to personalize, but personalizing the service aspects around them, such as delivery, does). AI can power the B2Me approach in supply chains in several key ways. First, AI can help supply chain leaders automate decision-making at scale. AI can analyze both structured and unstructured data and, thanks to that analysis, make real-time decisions, such as determining the shipping label for a package and assigning the appropriate carrier to an address. Automated decision-making at scale can help supply chain leaders focus on both performance and costs. AI can also power customer-aware logistics by integrating CRMs with logistics data in real time. The fusion of this information can enable supply chain leaders and their teams to access important details, such as a customer's lifetime value, their history of fraudulent activity and more. With that information, AI can make automated decisions accordingly. Moreover, supply chain leaders and their teams can make informed decisions when a situation calls for them to step in. Additionally, AI can provide supply chain teams with real-time alerts through a dynamic interface. So, if a package is rerouted due to traffic or weather delays, the company can get notified immediately so the team can take proper action. Then there's the unified data layer. With AI, supply chain leaders and their teams can access their data in one place, gaining a cohesive picture rather than searching for information across a scattered assortment of enterprise systems. AI can compile information from different parts of supply chains, connect the dots and offer that information. In turn, supply chain leaders will have greater visibility into their networks. Finally, AI—in this case, GenAI—can help supply chain leaders with simulations and strategic planning. Leaders can run multiple scenarios and see how different parts of the supply chain will react under certain conditions. With that information, they can create smoother workflows and stronger contingency plans. The Supply Chain Has A Soul—It's Time To Design For The Individual For decades, supply chains were engineered for efficiency, scale and predictability. They excelled at moving units, not understanding people. But the next frontier isn't about faster fulfillment or lower costs—it's about personalization at scale. It's about treating every customer not as a segment, but as a singularity. Here's how leaders should respond. Mass optimization alone will no longer cut it. In the B2Me world, designing for the average customer can lead to mediocrity. Leaders should re-architect their systems to respond to nuances such as shifting preferences, individual behaviors and contextual needs. Precision needs to replace standardization. And that requires a shift from planning for demand to sensing and shaping it in real time. You can't manage what you mismeasure. Traditional KPIs tend to reward internal performance, not external impact. It's time to elevate metrics that reveal how well your supply chain feels to the individual: delivery confidence, resolution friction, moments of surprise and delight. Because in a world of perfect substitutes, experience is the only moat. Data isn't just for dashboards—it's how your organization listens. The goal isn't more data; it's more meaning. Leaders must develop the muscles to turn messy, incomplete signals into actionable, human-centric decisions. What do your customer interactions really say about their trust, their frustration, their loyalty? Read between the rows. Most supply chains today are brittle—they follow instructions. Tomorrow's supply chains must be reflexive. They need to anticipate, learn, adapt. They must be less pipeline, more nervous system. A supply chain that can't course-correct midstream or personalize at the edge isn't just inefficient—it's invisible to the modern customer. The supply chain is no longer just infrastructure. It's experience architecture. If you're not designing it to serve individuals consistently and contextually in a personalized manner, then someone else will. Forbes Technology Council is an invitation-only community for world-class CIOs, CTOs and technology executives. Do I qualify?
Forbes
3 days ago
- Business
- Forbes
Angelic Intelligence: Designing Machines To Learn What It Means To Be Human
Shekar Natarajan is the founder and CEO of Nearly 20 years ago, I held my father's hand as I made the decision to take him off life support. No algorithm could guide that moment. No dashboard could absorb its cost. The question wasn't 'Can we?' but 'Should we?' In that moment, I understood that intelligence without morality is just mechanical cruelty. Later, as a supply chain leader, I watched logistics systems optimize efficiency by dehumanizing the very people who kept them running: warehouse workers denied bathroom breaks, drivers penalized for helping elderly customers. I saw the same pattern repeating at scale. We have built systems obsessed with process but indifferent to people. My mother's moral code showed me early on why human will and morality matters. So, I began to ask: What if technology could be more than rational? What if it could aspire to the best of us? At we are building what I call Angelic Intelligence, a framework that embeds a moral cortex in AI so machines don't just imitate human knowledge but honor human values. Our goal is not to make machines more intelligent but to make them more humane. 1. The Moral Cortex Layer (MCL) A programmable ethics engine that works like an AI's prefrontal cortex. It introduces ethical speed bumps when decisions have moral consequences, forcing systems to pause, reflect and justify their choices instead of rushing to the most efficient outcome. Imagine a routing algorithm suggesting a driver skip the last delivery to avoid overtime penalties. That delivery could be vital to a customer and not getting it could have grave consequences. The MCL interrupts the process, surfaces the decision for review and allows a human to approve the exception if compassion outweighs cost. Most systems default to binary decisions. The MCL keeps judgment in the loop, making sure the gray areas of human intent are respected and visible. 2. CareNet: Empathy As A Service A real-time framework that captures, verifies and amplifies acts of compassion across the system. CareNet is like a Fitbit for empathy, collecting peer-nominated stories of courage, care and discretion. A technician stays after hours to help an elderly customer reconnect their device. Instead of reducing this to a productivity penalty, CareNet records it as an act of care. The system recognizes and logs the context, not just the time spent. Empathy isn't a performance metric; it is a human experience. CareNet ensures acts of kindness don't disappear into the margins. When compassion is made visible, it becomes part of the culture rather than an exception to it. 3. Human Signal Intelligence (HSI) A learning model that captures outlier decisions where human judgment overrides the algorithm and produces a better outcome. These signals are decentralized, context-rich and often ignored by traditional AI. A delivery driver defies an optimized route to check on an isolated customer. Later, the customer's health emergency is discovered in time. HSI captures this deviation and learns that sometimes efficiency must yield to care. Most AI systems treat exceptions as noise. HSI treats them as evidence. Learning from outliers helps machines understand the moral heuristics that make humans irreplaceable. 4. The Ethical Memory Vault A secure repository that collects and tags acts of moral courage. The Vault serves as a story engine, feeding these examples back into training, onboarding and leadership development. When a warehouse worker breaks protocol to protect a colleague from harm, the story is recorded, tagged and shared across the organization. It becomes part of the collective memory that shapes future decisions. Cultures aren't built from compliance manuals. They are built from stories. The Vault ensures lessons in courage and empathy are never lost or forgotten. 5. Pause Protocol Interface A real-time flagging tool that gives frontline workers the power to challenge AI outputs without fear of retaliation. The interface allows anyone to pause, escalate or question decisions as they happen. A customer service agent receives an automated prompt to end a call quickly. Believing the customer is in distress, the agent hits the Pause Protocol to halt the script and escalate the issue to human review. People often stay silent because they don't feel safe raising concerns. The Pause Protocol restores agency, protects dissent and keeps the system grounded in human judgment. 6. Compassion As A KPI A redefinition of success metrics to reward compassion, context and care alongside speed and scale to help balance operational goals with moral outcomes. Instead of measuring only call duration or deliveries per hour, the system includes a compassion score based on peer recognition and customer impact. Small acts of kindness are weighted in performance reviews. If your operation can't survive compassion, it doesn't deserve to scale. Balancing efficiency with empathy ensures your culture stays human even as it grows. 7. Human-Centered Governance (HCG) A framework that keeps people as the final moral authority over automated systems. It includes role-based access, dynamic guardrails and what I call a Red Button Layer—an explicit right to override. A regional manager reviews an AI decision that would eliminate a crucial delivery route to save costs. Using governance tools, they pause the action, gather input from stakeholders and redesign the process to protect vulnerable customers. Machines don't govern morality. People do. HCG ensures that no matter how sophisticated AI becomes, it remains subordinate to human conscience. A 'Should We?' Angelic Intelligence is not trying to be a utopian ideal. Instead, it is a practical response to a world where efficiency threatens to outpace empathy. We don't need smarter machines that can beat humans at pattern recognition. We need wiser systems that remember why the patterns matter in the first place. Because intelligence without conscience is just speed. And in the race to build the future, I would rather be slow for the right reasons than fast for the wrong ones. Forbes Technology Council is an invitation-only community for world-class CIOs, CTOs and technology executives. Do I qualify?
Forbes
15-07-2025
- Automotive
- Forbes
The Cow, The Code And The Chaos: Why Logistics Needs Hybrid Intelligence
Shekar Natarajan is the founder and CEO of Some time ago, I was a passenger in a vehicle in rural India—a place where there is no rule of law regarding driving. Drivers didn't stay in neat lanes; some ended up on the side of the curb. Drivers could simply raise their right hand to signal a move to the left. They could raise both hands to signal they'd be driving straight. Somehow, in the cacophony and chaos, my driver was able to understand the intent of the other drivers and weave through them. But then appeared on the road a living creature whose intent my driver could not understand: a cow. Neither could the cow understand the intent of my driver. Yet in that moment, both had to coexist on that road. Coexistence, I believe, is necessary for success in all areas of our lives—rather than one or the other, we must find space for both. Technology is not an exception. In my view, for technology to truly help us succeed, we have to approach it through the lens of coexistence. In the supply chain world, AI can help future-proof the industry. In particular, I've observed that GenAI stands to create significant transformations. Yet, for optimal success, GenAI should be complemented by execution engines. The Limits Of GenAI Alone In The Supply Chain World Why is solely using GenAI not ideal in the supply chain world? GenAI behaves probabilistically; it is more predictive and speculative in nature. It exists to simulate, generate and explain. It can take data, uncover patterns and forecast possibilities accordingly. But the nature of the logistics industry is not probabilistic. Logistics isn't about what could happen—it's about what does happen. GenAI can only go so far in a non-probabilistic reality. For instance, a GenAI solution could analyze past data points and predict that if a retailer ships a package with a certain carrier in a certain city, there is a 90% chance the package will arrive on time. However, that prediction only holds weight in a digital world, which lacks the nuances and disruptions of the physical one. Severe weather, a port blockage, etc., can all cause shipping delays, regardless of the carrier. GenAI And Execution Engines: The Case For A Hybrid Approach By contrast, execution engines are deterministic software systems that choreograph and execute concrete actions. They exist to act, decide and choreograph. They revolve around intelligent action, which is the ability to take contextual actions in the physical world. However, acting alone, executive engines lack the flexibility to adapt in real time. Combined, GenAI and execution engines create a hybrid type of technology called execution engine optimizers. Execution engine optimizers fuse GenAI's reasoning with the capabilities of execution engines to take the most optimal actions in real time. So, GenAI essentially serves as an overlay engine that comprehends intent and simulates. The execution engine does the rest. Execution engine optimizers bind the two worlds together in an intelligent manner, transforming intent into outcomes. For example, a customer could enter a tracking number into a chatbox, and from there, GenAI in the backend could start its analysis to pinpoint who the consumer is, what their customer lifetime value is and where the package is. By overlaying different data points, such as the weather and traffic conditions, the technology could identify the root cause of the delivery failure (such as a delay or theft). GenAI could then determine the appropriate response to provide the consumer based on the situation. From there, the execution engine could automatically process both a carrier claim and check inventory availability, then offer the customer options such as reshipment or a refund. If the customer, say, opts for a reshipment, the execution engine could create a new shipping order in the system, generate a new tracking ID and send the customer an automated update with the new tracking ID. In short, these two technologies have different purposes. GenAI processes ambiguous, unstructured data. Execution engines analyze outputs and then execute accordingly. For the best results possible, both of these technologies need to coexist; they need to dance together. Power a supply chain system on GenAI alone, and you're banking on probabilities. Power one on just an execution engine, and you'll get actions without context. How Supply Chain Leaders Can Leverage The Hybrid Approach Supply chain leaders can leverage the hybrid technology, called execution engine optimizers, by taking several steps. First, supply chain leaders should identify their most significant friction points. Those are the problems that, if solved, will yield the most significant net positive results. Next, I recommend that supply chain leaders use GenAI for analysis and scenario planning. For instance, they could test to see which workflows will get impacted if certain variables change. This information will help them predict potential failures and categorize problems—from there, they should design specific, deterministic processes to address identified issues. At this point, they can take the outputs GenAI has provided them and start connecting them to execution logic (in other words, automated actions). As the generative AI engine and execution layer run, supply chain leaders should monitor how the systems are performing and create feedback loops that continuously optimize both the GenAI's and the execution engine's capabilities. On that road in India, at first, my driver honked and honked for the cow to move. The cow didn't budge. And then he, along with other people, started tapping the cow. The cow looked at everyone, then started walking away. Both parties—people and the cow—were able to create a more stable, streamlined environment. Supply chain leaders have the opportunity to stabilize and streamline supply chains if they put GenAI and execution engines in a symbiotic relationship and take action when needed to keep the duo running together harmoniously. Forbes Technology Council is an invitation-only community for world-class CIOs, CTOs and technology executives. Do I qualify?
Forbes
08-07-2025
- Forbes
The Code Of Her Convictions: What My Mother Taught Me About Building Moral AI Architecture
Shekar Natarajan is the founder and CEO of When I think about the most advanced system I've ever seen, I don't think of an algorithm. I think of my parents. Growing up in the slums of India, I learned from my father's hand how systems work. I learned from my mother's heart that no system, no matter how entrenched, works without the engine of humanity. My mother only finished 10th grade before she married my father to help raise her orphaned sisters, my aunts, as her own. She never studied engineering or computer science and had no formal training in logistics or AI, yet she built systems with nothing but will, compassion and quiet resolve. These systems shaped who I am and how I lead. One of those systems began with a rule. The Rule, Refusal And The Ring Where I grew up, there was a policy: No more than two children from the same family could attend the city's top school. My two older brothers were already enrolled. I was the third. According to the system, I didn't qualify—not because I lacked the ability, but because I was born third. My mother didn't fight the rule with outrage. She wore it down with persistence. Every morning, for nearly a year, she waited outside the education minister's office. She timed her arrivals with the daily mass schedule so she wouldn't miss him. She never raised her voice. She simply outlasted the system. Eventually, the headmaster relented. But when admission came, the tuition—30 rupees, barely 50 cents—was still out of reach. So she quietly removed the silver toe ring that symbolized her marriage and pawned it. That was how I got my seat in that classroom. Not only did she bend a system to make way for her child, she reassembled it with grace and grit. She didn't just endure the system. She made it work. And then she helped it work better. She hand-bound textbooks so they would last far beyond our family's needs. She turned scarcity into continuity and continually reused, repaired and repurposed. She didn't just endure the system. She made it work for people it wasn't built to serve. When The System Breaks, Humans Step In My mother's determination taught me about supply chain resilience long before I knew what the words supply chain and logistics meant. When I studied industrial engineering at Georgia Tech, I focused on learning, understanding and improving the global supply chain. What I learned most is that, although logistics seems to be about manufacturing and warehouses and trucks, it's actually a people-centric business. When software crashes or delivery networks buckle, systems don't save us. People do. • A warehouse associate reroutes pallets when the screen goes dark. • A planner reshuffles shipments at 3 a.m. to meet demand that the forecast missed. • A driver completes a route by memory when the app freezes. • A seasonal employee works late, so your package arrives in time for Christmas. It's human will that ensures continuity, not technology. The Risk We Run With AI There's no doubt that AI is an important and powerful tool that will help reshape the global supply chain. We celebrate AI's ability to optimize, predict and adapt, but we rarely ask: Can it care? Can it sacrifice? Can it pause when something just doesn't feel right? AI doesn't yet understand what my mother understood instinctively: that fairness isn't always about equality, it's about equity, and that logic has its limits. But that compassion belongs in decision-making. If we aren't intentional, AI will reduce human will to a variable—something to eliminate, rather than elevate. The same spirit that saved my education, that ensures your holiday package arrives on time, could be written out of the equation altogether. That's why we need a new kind of architecture for the age of machines. Angelic Intelligence (AIx) A Moral Operating System For The AI Era At we're building what I call the Angelic Intelligence framework: a set of principles and tools designed to ensure technology preserves what makes us human. 1. Moral Cortex Layer (MCL): A programmable ethics engine that evaluates system decisions against principles like fairness and harm reduction. It introduces pause points when high-stakes outcomes arise. 2. CareNet – Empathy-as-a-Service: A real-time API capturing acts of care, peer-nominated compassion and frontline discretion. Think of it as a Fitbit for empathy. 3. Human Signal Intelligence (HSI): A learning model trained on the very edge cases traditional systems ignore—times when a human overrides the logic and gets it right. 4. Ethical Memory Vault: A secure story engine that records acts of moral courage—feeding these back into leadership training, AI models and onboarding systems. 5. Compassion As A KPI: A shift in performance metrics to reward care, context and human judgment—not just speed and scale. 6. Pause Protocol Interface: A frontline tool that lets workers flag questionable AI outputs—anonymously, without fear of reprisal—restoring agency to the edge of the system. 7. Human-Centered Governance: Executable layers of company values that act like guardrails around automation. Machines don't govern morality—people do. This is the future of software design. The way we train it is the way it's going to behave. We must bring humanity into the equation. It's the moral code behind our code. Building Systems That Mirror The Best Of Us My mother's legacy wasn't her education or her job title; it was the will she encoded into the people around her. That's the real 'legacy code.' In the supply chain, we say people are our greatest asset. But the moment metrics falter, we turn to cost cuts, not care. We call humans inefficient. But humans are the only reason systems move at all. Culture doesn't live in KPIs. It lives in the planner who covers a shift. The associate who works through the storm. The mother who trades a wedding ring for tuition. If AI is to play a central role in our future, then it must learn from the quiet strength of those like her—people who do not optimize, but endure. If we want AI to lead the future, it must be guided by the same code that moved my mother—not just logic, but love. Forbes Business Council is the foremost growth and networking organization for business owners and leaders. Do I qualify?
USA Today
30-06-2025
- Business
- USA Today
Last Mile Solutions ("LMS") and Orchestro.AI Partner to Power Smarter E-Commerce Delivery Networks
Last Mile Solutions (LMS), a high-performance parcel carrier designed for e-commerce, and a logistics intelligence platform transforming delivery decision-making, today announced a strategic partnership to build smarter, more adaptive shipping infrastructure for the modern commerce economy. LMS operates a flexible, California-compliant logistics network serving high-growth sectors including meal kits, fresh produce, and traditional e-commerce. With consistent next-day service, LMS meets the rising demand for precision, speed, and regulatory alignment in complex urban, suburban, and rural environments across the U.S. Through this partnership, LMS will embed Orchestro's intelligence layer into its operations – enhancing network planning, market responsiveness, and decision velocity. Orchestro will also enable integration into third-party systems – including WMS, TMS, OMS, and CRM platforms – to streamline data exchange and execution across the supply chain. The platform's predictive capabilities allow LMS to adjust delivery operations based on live weather patterns, traffic disruptions, and localized risk signals such as theft or social unrest. These signals power near real-time communications with customers and help shippers reduce losses tied to delay, damage, and package theft. The partnership enables LMS delivery stations to move beyond static dashboards and interact directly with their operational data – using conversational AI agents and augmented intelligence to make faster, smarter decisions in real time. Whether resolving a delivery exception or rerouting a driver, these tools bring intelligence to the edge of the network, where timing and context matter most. 'LMS delivers where legacy systems fall short – in sectors that demand both speed and precision,' said Doug Schwartz, CEO and Co-Founder of Last Mile Solutions. 'This partnership gives us the visibility and intelligence we need to act decisively – while remaining committed to operational excellence and customer trust.' Together, the companies will deploy solutions that help LMS grow its customer base, optimize delivery performance, and streamline operations across high-volume markets. The partnership also supports LMS's expansion into a nationwide delivery network by orchestrating seamless collaboration with national and regional carrier partners – delivering consistent service, intelligence, and value at scale. 'We are architecting a new era of logistics – one where intelligence moves with the network,' said Shekar Natarajan, Founder and CEO of 'By partnering with carriers purpose-built for e-commerce, like LMS, we're transforming how delivery happens at the edge: dynamic, data-driven, and tuned for the speed of modern demand.' About Last Mile Solutions LMS is a California-based parcel carrier built for the demands of modern commerce. LMS Core: LMS's proprietary next-day delivery network powered by our professional, uniformed, background-checked drivers in commercial delivery vehicles – faster delivery, better service, lower cost. LMS Reach: LMS's network of carrier partners, complementing LMS's Core network to enable fast, cost-effective delivery to 100% of zip codes in the continental U.S. LMS Lift: Revolutionary middle mile platform, enabling 2-day delivery to over 70% of the US population at rates cheaper than traditional 5-8-day ground competition. LMS combines speed, regulatory compliance, and operational transparency across a flexible last mile delivery network. For media inquiries, please contact: Shekar Natarajan shekar@ About is a logistics intelligence platform that turns shipping data, carrier agreements, and fulfillment workflows into real-time decisions. Its AI-driven tools help carriers, shippers, and store networks plan smarter, operate faster, and adapt dynamically to an evolving delivery landscape. SOURCE: View the original press release on ACCESS Newswire
