AI Support
AI-powered support for CSRD reporting with AI-supported functions in every phase – from intelligent data completion through automatic text generation to automated plausibility checks. You choose the AI provider yourself and configure it per team with your own access – for more efficient ESG reporting with full data control.
AI GOVERNANCE
AI Under Your Control – Provider Configurable per Team
AI features support your ESG reporting across modules: materiality assessment (discovering IROs with AI from your own records, in all ten topics), data collection (emission factor suggestions, NACE code suggestions, material and waste classifiers), report drafting (report section creation and narrative assistance in all ten disclosure reports), plus targets and analysis (target, goal and measure formulation, species research, location sensitivity narratives and discovery of ecosystem dependencies). You choose the underlying AI provider yourself and configure it per team – with OpenAI, Langdock or an OpenAI-compatible model and your own access. This keeps you in control of where and how your data is processed.
- AI-supported functions across all phases of the CSRD value chain
- AI provider freely configurable per team (OpenAI, Langdock or OpenAI-compatible)
- Your own API access – you decide which provider processes your data
- Control over data disclosure and processing location
- Switch-off options as well as call and time limits per AI feature
- Suggestion functions: results are offered for adoption, never saved automatically
- Where authoritative reference lists exist (incl. critical raw materials and waste codes), the system classifies deterministically – the AI does not guess
- Three-level progress indicator during generation
DATA COMPLETION
AI-Powered Data Completion
The integrated AI accelerates data collection in your ESG reporting: emission factor suggestions in the greenhouse gas calculation, NACE code suggestions when creating organizations, and material and waste classifiers in the circular economy – each against stored reference lists and consistently offered as suggestions for adoption.
- Emission factor suggestions in the greenhouse gas calculation
- NACE code suggestions when creating organizations
- Material and waste classifiers in the circular economy (against stored reference lists)
TEXT GENERATION
Automatic Text Generation for Sustainability Reports
Report section creation and narrative assistance generate structured text drafts for all ten disclosure reports based on the collected data – in ESRS-compliant language. All generated texts can be individually customized and approved.
- Report section creation and narrative assistance in all ten disclosure reports
- ESRS-compliant report language
- The report language is the active interface language – AI narratives are created in this language, supported by the glossary
- Individual customization and approval
QUALITY ASSURANCE
Plausibility Checks and Quality Assurance
The completeness assessment per report shows the ratio of filled to required data points, the plausibility check against the previous period detects outliers with alert thresholds and a confirmation dialog, and the target forecast projects target achievement year by year with probability of achievement. This ensures data quality before your report goes to external audit.
- Completeness assessment per report from the ratio of filled to required data points
- Plausibility check against the previous period with alert thresholds
- Target forecast: year-by-year projection of target achievement with probability of achievement
- Rule-based consistency and balance checks – e.g. twelve consistency rules in the S1 demographics and the water balance check with period lock in E3
- Allocation plausibility check for supplier data that rejects a 100 % allocation as methodologically flawed
- Ensure data quality before external audit
RECOMMENDATIONS
AI Recommendations and Analyses
The AI supports you from setup to ongoing analysis: "Discover IROs with AI" reads what you have already recorded in preeco – company profile and sector, sites, products and markets, suppliers and existing IROs – and suggests up to ten impacts, risks and opportunities, each naming the records it is based on; scoring and the materiality decision stay with you. NACE code suggestions classify your industry based on the company description. Discover IROs with AI is available in all ten materiality assessments.
- Discover IROs with AI: up to ten suggestions from your own records, with their source and without scores
- NACE code suggestions based on the company description
FAQ
Frequently asked questions
Each team configures its own provider under integrations: OpenAI, Langdock or a freely definable compatible provider with its own endpoint address, each with its own model and access key. Without an active configuration, the software hides all AI triggers. You decide whether your team uses AI at all, and with which provider.
Yes. Each function can be switched off and has call and time limits. Every AI function is started with the same button, "Generate with AI"; a three-step progress indicator shows whether the request is being prepared, the content generated or the result saved. AI functions are activated per tenant, that is, per team.
No. The AI functions are designed throughout as suggestions: results are offered for acceptance and can be reviewed and edited first. In the materiality assessment, accepted suggestions count as unreviewed and must be checked before activation. The work gets done with AI and your team, and your team makes the decisions.
Where an authoritative reference list exists, the software assigns values deterministically. Critical raw materials are matched against the EU Critical Raw Materials Act, and the waste classifier picks waste codes exclusively from the 842 entries of the European List of Waste. Which disclosures in a report are still open is determined by the application itself, not by the AI.
Narrative sections are written in the active interface language, and materiality assessment suggestions in the team's working language. The interface is available in 24 languages. All sections of a report can be generated together with one button; if a generation fails, the section says so and offers a retry. Exports never contain an error note.
preeco | sustainability contains more than 30 AI-assisted functions in 13 function families. They include the discovery of impacts, risks and opportunities in all ten materiality assessments, report sections in all ten disclosure reports, target and action wording, NACE code and emission factor suggestions, species research and ecosystem dependencies, and material and waste classifiers.
Terms on this page, briefly explained
Experience AI-Powered Reporting
Schedule a free consultation and see the AI features of preeco | sustainability in action.