Take the
first step.

The ESG Doc Chat is an interactive tool that lets users ask questions directly against their own ESG and sustainability documents, annual reports, human rights policies, environmental data sheets, governance frameworks, and more — in real time. Users upload one or more documents and submit natural language questions; the AI generates answers grounded in the document content and provides source references for each response.

This is particularly useful for quickly locating specific data points (e.g. "What is our Scope 2 emissions figure for 2024?"), verifying whether a policy covers a particular topic, or preparing for rating agency queries. Multiple documents from the same company can be uploaded simultaneously. However, running cross-company assessments within the same Doc Chat session is not recommended, as mixing different organizations' documents may produce inaccurate or conflated results. For multi-company screening, Briink's Batch Screener tool is the appropriate workflow.

🚨 IMPORTANT: Running an assessment on multiple documents that do not pertain to the same company is not recommended, as it might lead to inaccurate and misleading results.

Briink helps teams prepare for ESG ratings, including CDP, EcoVadis, S&P Global CSA, MSCI, ISS ESG, and Sustainalytics, by automating the data collection, structuring, and questionnaire workflows that typically consume the most time in a ratings cycle.

The platform extracts relevant ESG data from your existing documents, sustainability reports, policies, certifications, audit reports, and prior submissions, and pre-fills rating questionnaires with structured, source-backed answers. Every response is linked to the exact document and page number it was drawn from, making it straightforward to verify and defend each answer to rating agencies.

Beyond pre-filling, Briink identifies gaps against specific rating requirements, showing teams exactly where evidence is missing or where responses are unlikely to meet the scoring threshold. For CDP specifically, Briink offers a native integration and live pre-scoring functionality that estimates your score at the question and module level before submission, so teams can prioritize improvements where they'll have the most impact on the final rating.

Because Briink stores and indexes your ESG document library centrally, data prepared for one rating (e.g. CDP) can be reused and mapped across others (e.g. EcoVadis, S&P), eliminating the duplication of effort that happens when teams treat each rating as a separate project. This cross-framework reuse is one of the most significant time savings Briink delivers for organizations managing multiple ratings simultaneously.

Yes. When an organization’s name is typed into Briink’s screener, a PDF version of the webpage is generated and processed automatically. Briink's Web Scanner can also crawl across multiple pages of a website to find and extract relevant ESG content, for example, pulling sustainability commitments, governance disclosures, or emissions data from a company's corporate responsibility section.

For teams that need to screen many companies at once, such as investor portfolios or supplier bases, the Batch Screenings feature allows multiple companies to be assessed simultaneously using their web presence and any publicly available documents. 

Briink supports over 85 languages. Users can upload documents in any supported language, write questions in their preferred language, and request responses in a different language if needed. For example, a German-language sustainability report can be uploaded and queried in English.

Language preferences can also be set at the account level via the "user instructions" feature in the ESG Questionnaire Assistant, ensuring consistent output language across a full disclosure workflow without needing to specify it for each question.

Full list of supported languages:

1. Albanian, Albania
2. Arabic, Arab World
3. Armenian, Armenia
4. Awadhi, India
5. Azerbaijani, Azerbaijan
6. Bashkir, Russia
7. Basque, Spain
8. Belarusian, Belarus
9. Bengali, Bangladesh
10. Bhojpuri, India
11. Bosnian, Bosnia and Herzegovina
12. Brazilian Portuguese, Brazil
13. Bulgarian, Bulgaria
14. Cantonese (Yue), China
15. Catalan, Spain
16. Chhattisgarhi, India
17. Chinese, China
18. Croatian, Croatia
19. Czech, Czech Republic
20. Danish, Denmark
21. Dogri, India
22. Dutch, Netherlands
23. English, United Kingdom
24. Estonian, Estonia
25. Faroese, Faroe Islands
26. Finnish, Finland
27. French, France
28. Galician, Spain
29. Georgian, Georgia
30. German, Germany
31. Greek, Greece
32. Gujarati, India
33. Haryanvi, India
34. Hindi, India
35. Hungarian, Hungary
36. Indonesian, Indonesia
37. Irish, Ireland
38. Italian, Italy
39. Japanese, Japan
40. Javanese, Indonesia
41. Kannada, India
42. Kashmiri, India
43. Kazakh, Kazakhstan
44. Konkani, India
45. Korean, South Korea
46. Kyrgyz, Kyrgyzstan
47. Latvian, Latvia
48. Lithuanian, Lithuania
49. Macedonian, North Macedonia
50. Maithili, India
51. Malay, Malaysia
52. Maltese, Malta
53. Mandarin, China
54. Mandarin Chinese, China
55. Marathi, India
56. Marwari, India
57. Min Nan, China
58. Moldovan, Moldova
59. Mongolian, Mongolia
60. Montenegrin, Montenegro
61. Nepali, Nepal
62. Norwegian, Norway
63. Oriya, India
64. Pashto, Afghanistan
65. Persian (Farsi), Iran
66. Polish, Poland
67. Portuguese, Portugal
68. Punjabi, India
69. Rajasthani, India
70. Romanian, Romania
71. Russian, Russia
72. Sanskrit, India
73. Santali, India
74. Serbian, Serbia
75. Sindhi, Pakistan
76. Sinhala, Sri Lanka
77. Slovak, Slovakia
78. Slovene, Slovenia
79. Slovenian, Slovenia
80. Ukrainian, Ukraine
81. Urdu, Pakistan
82. Uzbek, Uzbekistan
83. Vietnamese, Vietnam
84. Welsh, Wales
85. Wu, China

Setting a language in the ESG questionnaire assistant with user instructions

Another way to change the language of your outputs is to use the "user instructions" feature in the ESG questionnaire assistant.

This can be especially useful when you have a questionnaire in English, but need to generate answers in another language.

Please note, this feature is currently only available for Briink's ESG Questionnaire Assistant.

You can learn more about how user instructions work in the dedicated article.

‍

The only small clarification here is that currently Briink primarily does its model benchmarking and ensures our model quality with english language data sets so cannot determine with the same level of certainty the quality of responses in other languages. However, based on feedback from our clients and partners, Briink's performance and results remain strong in many other languages.

Briink is powered by a combination of proprietary ESG-trained technology and large foundation models (LLMs), including OpenAI's GPT family. The platform is model-agnostic by design, constantly evaluating new models, including open-source alternatives, to ensure the best performance for specific ESG tasks. Different models may be used for different subtasks: extraction, classification, summarization, and answer verification may each use the most appropriate model for that function.

For questions about model security, data handling, or AI safety, contact business@briink or fill our contact form. 

Briink applies a "right-sized" approach to AI model selection using lightweight models for simpler classification and extraction tasks, and more advanced models only when the complexity of the task requires it. This minimizes energy consumption per query without sacrificing output quality. 

Additionally, rather than training proprietary large models from scratch (which has a significant carbon footprint), Briink builds on existing foundation models enhanced with ESG-specific logic and domain knowledge.

The platform is hosted on low-carbon cloud infrastructure, and all AI systems are governed by a Responsible AI Use Policy.