Skip to content
This is how we bring AI and Digital Product Passports together

When AI makes decisions about products, product knowledge is the foundation.

The Autonomous Expert Agent by Narravero connects product knowledge with up-to-date regulatory knowledge. It shows what in the portfolio is affected, what is missing and what needs to be decided.

The concept

An autonomous agent clarifies for every product what needs to happen now.

The Autonomous Expert Agent for Product Compliance takes over time-consuming groundwork: it continuously matches new and changed regulatory requirements against the product portfolio and shows affected products and markets, missing information and open decisions. Results approved by the team are used for global product compliance and transferred directly into Digital Product Passports.

NARRAVERO’S AGENTS

Highly specialised colleagues who support you and your team continuously and intelligently.

How an Autonomous Expert Agent works with your team.

Specialised in product compliance – with a clear mandate and defined approvals.

  • Takes Initiative

    The agent doesn't wait for the next prompt. It acts on its own.

    It spots data gaps, contradictions and need for action and prepares the next steps independently.

  • Always on the Job

    Compliance isn't a one-off check. Neither is the agent's work.

    It connects product knowledge continuously with curated regulatory knowledge and keeps track of changes.

  • Specialist by Design

    Not a generalist. Product compliance is its job.

    The agent connects your product knowledge with Narravero's specialised DPP and compliance expertise.

Colleague looking upwards, wondering
You're wondering: 20,000 products. One new regulation. Which ones are affected? Ask your new colleague.
Product Knowledge First

Rules become tasks across the product portfolio.

New requirements affect products, materials, supply chains and markets. The information needed is spread across ERP, PIM and PLM systems, documents and supplier data. The agent connects this product data with continuously curated regulatory knowledge and takes company-specific standards into account. Across the entire portfolio, it shows:

Technician holding an e-bike battery in the workshop
  • which products and markets are affected,
  • which information and proofs are missing,
  • where sources contradict each other,
  • which deadlines and decisions follow from this.
Example
Compliance Workspace
Your portfolio Carbon footprint Repairability Proof of origin Status
E-Bike-Akku410 units 6 × proof missingverifiednot affectedin review
Aluminium frame2,100 units not affectedContradictionnot affectedin review
Saddle860 units not affectedverifiedReply missingin review
Action needed verified not affected
Expert Agent

“I emailed the supplier yesterday. I'll follow up tomorrow if there's no reply.”

Control by Design

In-house sovereignty over data, access & processing.

The agent works within the sovereignty and protection framework defined by the company. The company decides

  • which data and systems the agent may access,
  • which information may be processed in which environment,
  • which data locations and external services are permitted.

Within this framework, the agent complies with every rule – for every task, at every level.

You set the boundary

Your data stays inside

Four access levels, from the workstation to the external service. You decide how far each type of data may go.

Example
Sovereignty and protection framework Defined by the company
Fully protected Increasingly open
  1. Level 1 Workstation nothing leaves the building Example: Formulations, bills of materials
  2. Level 2 Data centre your hardware, your administration Supplier contracts, prices
  3. Level 3 Controlled cloud isolated, fixed access rules Aggregated metrics
  4. Level 4 External service approved tasks only Public legal texts
Expert Agent

“Your contracts never leave the building. Public legislation I check externally.”

Model Orchestration

Technological freedom of choice in models, tools and computing power.

Compliance expert working at two monitors with a process diagram

The task determines the choice.

A general-purpose AI model knows neither a company's products nor its processes. The agent connects product knowledge from approved systems with the regulatory context.

For each task, it selects and combines models, tools and computing power – based on quality, sensitivity, speed, cost and control.

The value lies not in any single model, but in the intelligent selection and orchestration of the right tools within company guidelines.

Example
Compliance Workspace · Orchestration
Task Model Location Criterion
Material analysisBattery cell chemistry Specialist model Data centre Sensitivity
SummarySupplier report Fast model Controlled cloud Speed
Legal textsEU Battery Regulation Large language model External service Quality
Expert Agent

“Where speed matters, I use a fast model. Where precision matters, a specialised one.”

Global by Default

Global product compliance & Digital Product Passports.

Global thinking, regional execution: The same product knowledge can be used for Europe, China, Brazil, the USA and other markets — connected with the requirements that apply there. The agent prepares data gaps, conflicts and open decisions for expert approval:

Whatever is approved goes straight into the DPP process.

E-bike battery with QR code being inserted
The same product knowledge

Regionally connected, globally rolled out.

The Narravero Platform transfers it directly into the DPP – always up to date.

Example
Compliance Workspace · Markets
One product context
Action needed In review Ready for approval
Product

E-Bike-Akku 630 Wh

Cells · Material · Origin · Repair
  • EU Ready for approval
  • China 2 proofs missing
  • USA In review
  • Brazil In review
Expert Agent

“I've prepared the EU product passport for approval. China is still missing two material proofs – I've contacted the supplier.”

DPP process

Approved product knowledge, straight into the DPP.

The Narravero Platform transfers the approved information for the European market into the ongoing DPP process – for publishing, updating and rolling out across products and digital touchpoints.

Publish · Update · Roll out

01Ongoing DPP process

  1. Step 1Your team approves
    E-bike battery
    E-Bike-Akku 630 Wh Expert-approved
    Product knowledgeVersion 04
    Target marketEurope
  2. Step 2Narravero Platform

    Transfers the approved information into the ongoing DPP process.

    Synced · Version 04
  3. Step 3Digital Product Passport
    EU
    E-Bike-Akku 630 WhLive
    for the European market
    Cells
    Material
    Origin
    Repair

02What happens next

  • Publish DPP content is made visible.
  • Update Approved changes flow in continuously.
  • Roll out Products and digital touchpoints are reached everywhere.
The difference

Generic AI tools compared with the Autonomous Expert Agent for Product Compliance

Generic AI tools

  • Training data of unknown currency
  • Data leaves the company
  • One vendor, fixed lock-in
  • A text answer without traceability

Autonomous Expert Agent

  • Curated regulatory knowledge from standardisation
  • Data stays within the chosen framework
  • Freely orchestrated models
  • Traceable result in the product passport
Corporate AI sovereignty
  • Less vendor lock-in
  • Controlled AI costs
  • Auditable governance
  • Technological resilience
  • Faster response

Those who control their own product data and domain knowledge stay economically and technologically in control.

Book a consultation