Kovant
Kovant provides an agentic platform that deploys teams of autonomous AI agents to run enterprise supply-chain operations end-to-end, covering procurement, inventory, fulfillment, and demand. The platform operates within the customer's own cloud tenant on existing systems, with a reasoning core deciding next actions and a policy core validating every action against customer rules. It can go live in weeks, starting with one lane, and escalates only true exceptions to named human owners.
- AI Agents
- Logistics & Supply Chain
- Artificial Intelligence
- Manufacturing / Industry 4.0
- Enterprise Software
- Software Only
- Industrial Automation
Funding
Raised to date
€1.5MRaised to date based on public sources. This may differ from the amount the company actually raised and is based only on what is publicly available on the internet.
Est. $1.6Macross 1 round
AEMSH€1.5M · Pre-Seed
Announced Nov 2025 · from article date
Disclosed cumulative: $1.6M
Founders
Founder details are not available yet.
Product
Problem
Enterprise supply-chain operations remain heavily dependent on manual processes and brittle, scripted workflow automation that breaks when exceptions occur. Unstructured inputs like emails and documents, combined with siloed systems, make it difficult for organizations to scale operations efficiently and respond dynamically to change. Inventory levels are the result of thousands of weekly decisions that human teams cannot fully reach, leading to excess working capital or lost sales.
Solution
Kovant provides an agentic platform that deploys teams of autonomous AI agents as digital workers to run enterprise supply-chain operations end-to-end. The platform uses role-based execution, where agents are defined by job descriptions, policies, and escalation rules, rather than rigid workflows. Kovant agents hold a book of work in procurement, inventory, fulfillment, and demand, and run it end to end within the customer's own cloud tenant on existing systems. A reasoning core decides what should happen next, and a policy core validates every action against customer rules before it reaches company systems. Agents handle variation, communicate through familiar channels like Microsoft Teams, Slack, and email, and escalate only true exceptions to named human owners. The platform includes built-in governance, audit trails, and human oversight, and can be deployed in weeks without development or rip-and-replace, enabling immediate operational capacity. The first operation goes live in weeks, starting with one lane and scaling from there.
Target Audience
Primary customers are enterprise organizations in industrial, manufacturing, and supply-chain-intensive sectors, including procurement, inventory, production, and aftermarket service operations, seeking to modernize their operational core with autonomous AI execution.
Features
- Twelve foundation agents combine into cross-functional teams that plan, communicate, and execute under human oversight, scaling across operations and adapting to exceptions.
- Constraint-aware capacity modeling evaluates machine capabilities, labor availability, shift schedules, and material supply simultaneously to generate realistic capacity plans.
- Demand-driven scenario planning analyzes historical production data, forward order books, and market signals to simulate capacity scenarios and identify emerging constraints.
- Condition-based maintenance scheduling generates rolling maintenance plans from asset health signals, OEM intervals, and production commitments, automatically creating work orders.
- Spare parts planning projects requirements from maintenance schedules and supplier lead times, triggering procurement before stock-outs occur.
- Agents track utilization rates, work-in-progress levels, and lead times across lines and workcenters, flagging queue imbalances and capacity variances proactively.
- Agents commit delivery dates against real availability, allocate stock across competing orders, and escalate orders no longer achievable before they reach the customer.
- Model-agnostic architecture supports Small Language Models (SLMs) to reduce hallucination risk, with ISO 27001-certified security and optional deployment in the customer's own cloud tenant.
- Every action an agent takes is logged with its inputs, its reasoning, and the resulting record, with permissions configured per lane and actions outside those limits escalated to a named owner.
- Accessible through Microsoft Teams, Slack, and email, with a central Agent Hub (Concierge) for interaction and management, plus Mission Control for governance.