Command every AI agent across the osapiens HUB
Orchestrate and monitor AI agents across every osapiens HUB solution from one place. You set the guardrails and make decisions. The Agents handle the repetetive work, while you focus on your business.
Impact at scale
One connected context for Every Agent
Works across the osapiens portfolio, not just one solution
Agents sit in the Agentic Core, between Data Layer and Solution Suites, checking, monitoring, and enhancing continuously during use. This gives them the ability to act across solution suites, including Supplier Intelligence, Reporting & Disclosures, Corporate Carbon Footprint, Traceability, and Maintenance.
Benefits from unified data the moment it enters the HUB
Data enters a single unified structure as soon as it reaches the HUB, whether from a UI input or a system integration. A global knowledge graph defines each parameter and its meaning once, and that definition becomes available to every agent in the customer's tenant, making results even better.
Agents get smarter by tapping into our internal Knowledge Graph
Instructions and mappings created by osapiens combine a customer's own data, solution context, and domain knowledge into a single context for each agent. An internal abbreviation that would otherwise confuse a generic model gets recognized correctly, alongside its full meaning, because the knowledge graph carries that context into every relevant task.
A growing library of ready-to-run Agents
New agents ship release after release, and the library expands without any work required from the customer. For what goes beyond the standards provided, your own agents help you cover what suits your processes best.
Built-in AI capabilities
Every agent in the osapiens Agent Factory draws on a shared set of capabilities, including retrieval-augmented generation for external context, named entity recognition, speech-to-text, text rewriting and summarizing, translation, language detection, intent recognition, structured data extraction, and role-based chat, with new capabilities added as the Agent Factory evolves.
Choose the model behind each agent
Agents by default use open source models hosted inside the HUB, making sure your data stays inside the HUB. Where required, switching to commercial models can be triggered as an active decision. Application logic never touches a raw model directly; it calls the Agent Factory gateway, which pre-processes data and constructs context before a model is called.
Built-In Intelligence for Every Solution Suite
Domain intelligence, built in
Agents deployed in the osapiens HUB already understand the processes in a customer's solution space, supply chain, and data from day one. No custom training, no prompt engineering, and no integration project is required.
Connect to data you can trust
Agents connect directly to structured, verified HUB data in real time. No data preparation, no data pipeline risk, and no stale inputs skew the output.
Govern every agent by design
Explainability, human sign-off, audit trails, and data sovereignty are built into every agent from the start, not added afterward.
osapiens AI Agents
Explore by suite
Agentic support for Supplier Intelligence
Agents that keep every business partner record complete, current and trustworthy, from onboarding and questionnaires through risk trends to case resolution.
Learn more about Supplier Intelligence-
Business Partner Deduplication Agent
Identifies likely duplicate supplier records across spelling variations and incomplete data, clustering results for consolidation. -
Case Enrichment Agent
Enriches new compliance cases with ranked assignees, resolution steps, and risk assessment, resolving clear-cut cases without human effort. -
Case Management Agent
Collects and validates supplier information autonomously, delivering a complete summary for human decision-making. -
Questionnaire Auto-Fill Agent
Extracts and pre-fills questionnaire fields from uploaded documents, removing manual data entry and preventing supplier fatigue. -
Record Validation Agent
Checks every Supplier Intelligence record against a configurable quality checklist automatically. -
Task Assignment Agent
Analyzes new or unassigned tasks and suggests the most appropriate owner with a one-line rationale. -
Business Partner Onboarding Agent
Automates the full business partner onboarding lifecycle: linking, enriching, certificate scanning, and dispatching questionnaires, with a summary delivered by email. -
Supplier Risk Trend Agent
Turns supplier risk score data into a ranked view of the biggest movements, removing manual analysis.
Agentic support for Product Compliance
Agents that keep product, origin and process data clean and traceable across every regulation you report against.
Learn more about Product Compliance-
Address Normalization Agent
Standardizes country codes, city names, and state descriptors so origin records stay consistent and ready for analysis. -
Supply Chain Monitor
Detects origin deviations, unusual patterns, and unknown process steps across a product's supply chain at scale. -
Origin Certificate Assistant
Extracts origin information from uploaded certificates, images, and delivery notes, and prepares it for event reporting. -
Code Management Forecaster
Predicts production code usage for products and machines ahead of lead time, so code stock is ready when production is planned.
Agentic support for Carbon Management
Agents that take the manual work out of emission data, mapping factors, reviewing inputs and flagging what looks wrong before it reaches a report.
Learn more about Carbon Management-
EF Mapping Agent - Procurement Emissions
Matches emission factors from Ecoinvent and Climatiq to procurement group context, surfacing each recommendation for human review. -
Data Review Assistant
Flags year-over-year spikes, missing entry types, and unit inconsistencies before they become reporting errors. -
Code Management Forecaster
Predicts production code usage for products and machines ahead of lead time, so code stock is ready when production is planned.
Agentic support for Reporting & Disclosure
Agents that turn existing evidence, targets, and policies into structured disclosure content, reuse answers across standards, and draft contributor responses for team review.
Learn more about Reporting & Disclosure-
Cross-Standard Report Agent
Maps and adapts report content across standards, carrying relevant answers between frameworks to remove duplicate writing effort. -
Disclosure Task Agent
Turns defined actions, targets, and policies into structured disclosure content automatically. -
Evidence to Report Agent
Turns existing certificates, policies, and reports into structured disclosure content -
Strategy Disclosure Agent
Drafts task responses for each contributor from data already in the system, so teams focus on review rather than drafting.
Agentic support for Maintenance & Repairs
Agents that turn incoming requests into complete work orders, plan and assign them, guide technicians through execution, and close clean records while flagging exceptions.
Learn more about Maintenance & Repairs-
Maintenance Co-Pilot - Execution Agent
Guides technicians through work orders step by step in chat, answering questions from manuals and history while capturing parts, time, and notes automatically. -
Maintenance Front-Desk
Turns information captured in any format into a clean, complete note ready to save without manual tidying. -
Maintenance Notetaker
Captures maintenance issues from email and other channels, asks follow-up questions, and creates a complete ticket without form filling. -
Work Order Dispatcher
Plans and assigns work orders continuously by matching skills, availability, and location, replanning when breakdowns or absences disrupt the schedule. -
Work Order Drafter
Converts every incoming ticket into a fully enriched work order, drawing on historical data to cut planning time. -
Work Order Reviewer
Closes clean work order records automatically and surfaces only the ones that need attention.
You Want an Assistant, Not an Autopilot
Agents handle repetitive, high-confidence work and improve as they learn through use. Judgment, exceptions, and accountability stay with the customer's team. Compliance, ethics, and governance are not features added on top of agentic AI; they are the foundation it runs on.
Structure and customization — Agents ship ready to run and can be reconfigured by the customer. Misconfiguration is effectively eliminated, as each agent is inherently aligned with the capabilities of its solution and built by combining nodes into an end-to-end process within that solution's context.
Dynamic guardrails — Customers configure how agents respond in specific situations, including stopping interaction if a conversation turns offensive or handing over to a human when a defined trigger is reached.
Monitoring and human-in-the-loop — Agents can request human input at any point. The context of a running agent can be inspected at any time, every agent flow is fully auditable, and an agent's position in its flow stays visible.
Keep Every Agent Accountable to Your Team
Every agent stays configurable, auditable, and answerable to the guardrails a customer sets. Judgment and exceptions stay with the team; the Agent Factory handles the repetitive work underneath it.