Technology

Pedestal AI Raises $4.7 Million to Use AI Agents in Supply Chains

Martin HollowayPublished 20m ago3 min readBased on 1 source
Reading level
Pedestal AI Raises $4.7 Million to Use AI Agents in Supply Chains
Image by cookieone from Pixabay

Pedestal AI emerged from stealth on November 17, 2026, with $4.7 million in seed funding led by Eclipse. RRE, Converge VC, Timescale VC and VitalStage also took part in the round, according to TechCrunch.

Raj Aggarwal is the company's CEO. Pedestal AI says its goal is to use agentic AI, software that carries out multi-step tasks with limited human direction, to make supply chains faster and more reliable.

The company plans to put the new capital into three areas: expanding its product and engineering team, growing its marketing organization, and supporting additional deployments. These are plans, and the funding is intended to pay for them.

The one customer result cited so far involves a major furniture company. Pedestal AI reports a 20% reduction in that company's missed deliveries after deploying agents to handle work that drivers were supposed to do. The mechanics are specific. Thirty minutes before a delivery, Pedestal's system calls the customer to collect special instructions. If the call does not connect, a text goes out from a separate system. The result is a workflow with a primary channel, a fallback channel and a defined information-gathering step, all completed before the truck arrives.

In my view, that workflow is a useful lens on what agentic AI is being asked to do in logistics. The task is narrow and time-bound. It has a defined input (special delivery instructions), a deadline (the 30-minute window) and an explicit failure path (the secondary text). Those properties make it measurable in a way broad productivity claims rarely are. A missed delivery is also an unambiguous failure, typically meaning a rescheduled attempt, extra handling and an unhappy customer, so a 20% reduction is easier to interpret than a general efficiency figure. It is still one customer's result. A seed-stage vendor will need to show the number holds across other deployments before it carries weight beyond this account.

The broader context here is that supply chain software has long been stronger at planning, tracking and optimization than at the final stretch, where someone has to be reached, asked a question and given a time. Much of that work has depended on drivers, dispatchers and customer service staff making phone calls and manual follow-ups. Pedestal's proposition is that software agents can take over parts of that coordination and do it more consistently than a queue of human tasks. Whether that holds at scale will depend on how the agent handles unexpected replies, how exceptions are escalated to people, and how the system's actions are audited.

The seed round also places Pedestal in a crowded field of companies applying large language models to operational workflows. Its early focus on a discrete, customer-facing task, rather than a general assistant layered over existing systems, is a narrower ambition, and narrower ambitions are often easier to verify.

In my view, if the approach holds, the gain is less about any single feature than about moving routine coordination out of the driver's queue and into software that can retry, escalate and log what it did. That is unglamorous work, and it is where agentic systems will either earn their place or fall short. Pedestal now has its funding and its first published result. The next useful data point will be how the same workflow performs with customers that are not furniture retailers.