In today’s fast-paced world, staying ahead of the curve requires more than just operational efficiency, it demands we boldly embrace innovation too. With AI embedded across supply chain functions, from forecasting to fulfilment, Sanjeev Siotia, CTO, Manhattan Associates explores how Agentic AI is shaping the future of supply chain commerce.

Most people by now are familiar with Generative AI (Gen AI) tools like ChatGPT that respond to prompts with natural-language text, but each prompt is treated as a standalone interaction, even if it looks like an ongoing conversation. Rather than simply generating responses, Agentic AI operates a ‘think, see, do’ approach. The think component is powered by large language models, like those used in Gen AI, while the see aspect refers to the system’s ability to observe and analyse its environment. But it’s the do element where the true power lies, delivering agents that can take actions, execute tasks, and even make decisions based on contextual knowledge. Simply put, Agentic AI isn’t just about responding to queries, it’s about enabling autonomous, task oriented behaviour that mirrors and complements how humans work.

AI is not a replacement

While AI will continue to play a critical role, it isn’t replacing traditional supply chain mathematics or people. Rather, it complements both. Established algorithms for inventory management, forecasting, and optimisation are still relevant. The value of Agentic AI is its ability to orchestrate tools, choosing the right one for the right job at the right time. It also redefines how users interact with software. Instead of clicking through complicated interfaces, users can simply ask, ‘Where’s my shipment?’ and receive a clear, actionable answer. Natural language interfaces reduce the need for specialised training, making systems more accessible to a broader workforce.

An agentic future

Not all systems are created equal when it comes to AI integration. For almost a decade now, Manhattan has invested in building a robust, API first architecture that allows and enables agents not just to ‘see’ the data but to ‘take actions’ based on it. Acting on data isn’t just about insight though. It’s about execution. Whether that’s adjusting shipping plans, notifying vendors, or triggering alerts, agents need safe, validated access to core systems. This is where Manhattan’s Active platform gives us the edge. It’s designed to let agents operate within guardrails, ensuring data integrity and security while unlocking efficiencies at scale.

To make it easier for teams to get started, we have introduced Manhattan Agent Foundry™, a toolkit for building, testing, and deploying agents. Users can start from templates, modify existing agents, or even use AI to build new ones - yes, that’s right, agents that can build agents! The system’s structure prevents errors, so users can experiment safely and learn in real-time.

AI is a clear technological and societal inflection point, much like the internet, smartphones, or cloud computing before it.

The inflection point

AI isn’t just a passing trend, it’s a clear technological and societal inflection point, much like the Internet, smartphones, or cloud computing before it. Whether we like it or not, AI is becoming an inextricable part of our professional and personal lives, so the best thing we can do is to start now, embrace it, experiment with it, learn and build. AI is here to stay and it’s not just enhancing supply chain commerce, it’s redefining it.

Sanjeev Siotia, CTO at Manhattan


Wilt u een fysiek exemplaar van het magazine ontvangen? Stuur een mail naar ce@manh.com o.v.v. Commerce Trends 2025, en vermeld bedrijfsnaam, uw naam, adres en woonplaats. Wij sturen u dan een exemplaar toe.