Automated Onboarding for Energy Communities

One of our clients in renewable energy needed to scale customer onboarding without growing the operations team. An AI-powered onboarding pipeline now takes prospects from utility bill upload to signed contract in minutes. With automated invoice extraction, community matching, and contract generation, the same team processes around 300 contracts per month, 24/7.

client:
implementation time:
3 Months
Technologies:
Gen-OS
industry:
Energy & Utilities
team in this project:
Pedro Nogueira
Project Manager
Tiago Mateus
Data Scientist

We operationalize data to deliver measurable impact

300
quotes generated per month
93%
invoice extraction accuracy
10-20s
processing time per document‍
Data measured within the first month in production

The Opportunity

Greenvolt Comunidades develops and operates renewable energy communities, enabling households and businesses to share locally produced clean energy and benefit from lower electricity costs. The company manages the full lifecycle from prospect acquisition to community activation. Onboarding into an energy community required a fully manual flow: reading utility bills, checking community availability, calculating savings, and generating contracts.

Manual, multi-step onboarding

Operators had to extract consumption data, match locations to communities, calculate savings, and prepare contracts case by case.

Heterogeneous invoice formats

Utility bills arrived in different layouts and from multiple providers, making structured extraction slow and error-prone.

Complex community availability logic

States like active, under construction, oversubscribed, and waitlisted each required different rules and customer journeys.

Capacity and timing constraints

Every conversion depended on operator availability, and delays increased the risk of losing prospects before contract signature.

The Solution

Using GenOS Supervisor, we designed an end-to-end onboarding pipeline embedded directly in the website.

A prospect uploads a utility bill. From that point, Supervisor orchestrates the flow: extracting more than 30 fields from any bill format, applying confidence scoring from the start, and routing each case accordingly. High-confidence extractions proceed automatically. Low-confidence cases are routed to a human reviewer before a proposal is generated, keeping accuracy under human control where it matters.

Downstream, contract generation is handled from templates, with electronic signature via DocuSign and IBAN and company certificate validation for corporate clients. In parallel, geospatial matching identifies the nearest available energy community and calculates a personalised savings and CO₂ reduction proposal. When no community is immediately available, the prospect joins a waiting list and is notified automatically when a match opens.

Every step is tracked in Microsoft Dynamics 365. If a prospect drops before signing, Supervisor triggers an automated follow-up with the proposal and a link to resume, closing the loop between AI, process, and commercial outcomes.

Underneath all is the GenOS Platform: every action and decision is logged and traceable, and the entire onboarding funnel can be monitored, tuned, and extended from real production data as products and rules evolve.

The Impact

The onboarding pipeline now processes around 300 contracts per month with the same team, running 24 hours a day instead of being constrained by office hours and manual capacity.

Invoice extraction accuracy reaches about 93%, with most documents processed in seconds. High-confidence cases flow straight through to contract, while edge cases are escalated to human review without blocking the rest of the pipeline.

By treating onboarding as an AI-orchestrated process rather than a sequence of manual tasks, the client gained a scalable, auditable foundation for growing energy communities and a blueprint for extending automation to future products.

300
quotes generated per month
93%
invoice extraction accuracy
10-20s
processing time per document‍

A word from our customers

Real enterprises solving real problems with AI systems built for reliability, transparency, and scale.

"Lorem ipsum dolor ementum tristique. Duis cursus, mi quis viverra ornare."
Generic placeholder image
Name Surname
Position, Company name
"Lorem ipsum dolor sit amet, consectetur aros elementum tristique. Duis cursus, mi quis viverra ornare."
Generic placeholder image
Name Surname
Position, Company name

"From day one, the DareData team earned our trust through outstanding communication and responsiveness."

Generic placeholder image
Head of Al Tech Lab @ Euronext

”We were very pleased with the training. The materials were adjusted to our needs and, in the end, we could take home some ideas that we could apply to our business.

Generic placeholder image
Data Coordinator @ Worten

“DareData Engineering has the resilience to make the effort in improving our development and production processes.”

Generic placeholder image
Lead Data Manager @ NOS Comunicações

"Their ability to bring clarity to the application of models in practice is amazing."

Generic placeholder image
Revenue & Margin Growth Manager @ Heineken
TRUSTED BY THE WORLDS LARGEST ENTERPRISES