ESG questionnaire response from one evidence library — every answer cited, checked and approved by its owner
Every answer drawn from one evidence library, cited and owner-approved.
See one case, screen by screen ↓Software that answers the sustainability questionnaires a company receives — from ESG rating providers, disclosure platforms, customers, tenders and investors — from one approved source. ESG Questionnaire Response keeps one approved answer per topic in an evidence library, fills each new questionnaire from it with a citation for every statement, and sends the rest to the people who own the answers.
How to choose AI solutions for life sciences →
Updated
A drug maker answers more than ten sustainability questionnaires a year: ESG rating providers, a climate disclosure platform, hospital and pharmacy customers, national medicines tenders, group purchasing organisations and investors. They ask the same twenty-odd things — Scope 1 and 2 emissions, science-based targets, water stress at API sites, the supplier code, the human rights policy, access to medicines — in different words, numbering and formats, each with its own deadline. The ESG reporting lead rebuilds every one by hand, chasing EHS, procurement, HR and legal for the same figures and documents.
The risk is the answer nobody re-checked. After a site sale the GHG inventory is restated, but the preliminary emissions figure has already gone to a hospital customer and is copied into the next draft. A 2024 “carbon neutral since 2022” answer is reused after the offsets ended, although green-claims rules require any such claim to be current and substantiated. A human rights policy past its review date is attached to three questionnaires without anyone noticing.
Estimated split of the 18-workday survey baseline for one questionnaire, by hand and with the solution.
Six agents read each questionnaire, fill it from the library, check every figure and claim and route the rest; section owners approve every answer.
Northwind Bio, 7 October. Six questionnaires are open — two ESG ratings, a hospital group, a national medicines tender, a group purchasing organisation and an investor — and a restated emissions figure is still out in the world. Here is what happens next, screen by screen, in the working solution.
Priya's coverage matrix lists 27 topics down and the 6 open questionnaires across: 40 answers approved, 19 drafted to review, 7 flagged, 1 gap. Scope 1 and 2 emissions is amber in two of them — the Clearwater ESG Ratings data review and the Brightwater Hospitals supplier questionnaire both carry 61,400 tCO2e. A new file waits at the top: the Meridian Group Purchasing ESG supplier RFI, forwarded by Sam Patel, 18 questions read and mapped to 7 topics, due 12 November.
“Uses 61,400 tCO2e — the library holds 58,920.”
The agents fill the Meridian column cell by cell: 12 approved answers reused, 3 drafted from cited passages — 83% of the RFI. The rest go to their owners. The carbon-neutral question goes to Grace Liu in Legal, and two questions the library has never answered — product carbon footprint and biodiversity near protected areas — go to Omar Haddad.
“Nothing in the library answers this yet. The facts register holds company-level emissions only. Product carbon footprints need a calculation from the EHS emissions data. Routed to Omar Haddad.”
The library answer for B1 is 58,920 tCO2e for 2025 — Scope 1 21,480, Scope 2 market-based 37,440 — with limited assurance, restated in June 2026 after the sale of a manufacturing site, in line with the GHG Protocol, and three sources. It is asked in 12 questionnaires. Priya owns Climate & energy, so she clicks “Update everywhere”: Clearwater DP-02 and Brightwater B.1 now read “Updated from the library today”.
“The assured figure goes into Clearwater and Brightwater, and a correction note is drafted for the submitted Halden Health answer.”
Halden Health received 61,400 tCO2e in March, so their submitted answer is never changed silently. The answer drafter writes a correction note from the facts register: the site sale, the base-year recalculation, the assured 58,920 tCO2e, and page 22 of the 2025 Sustainability Report. Sam Patel sends it as the account lead.
“Drafted by the Answer drafter from the facts register. Sam Patel sends it as the account lead.”
In the Brightwater questionnaire, B.1 is approved by Priya Raman, with its three sources open beside it: the restated GHG inventory, the limited assurance statement from Corrigan Vale Assurance LLP and the 2025 Sustainability Report, each at the passage that supports it. The trace shows how it was built: questionnaire reader 0.4 s, evidence matcher 1.1 s, answer drafter 2.3 s, consistency and claim checker 0.8 s.
“Updated from the library. The assured 58,920 tCO2e replaced the preliminary 61,400.”
The draft for “Are your operations or products carbon neutral?” reused a 2024 Ostara Pharmacy Group answer. The offset programme ended on 1 January 2026, and the climate transition plan says the company does not describe its operations or products as carbon neutral. The checker strikes the old text and offers approved claim wording from Legal, with the 34% reduction since 2021 and the net-zero 2045 aim, each cited.
“Risky claim: ‘carbon neutral’ is not supported by current evidence.”
The Human Rights Policy was approved in September 2023 and says it is reviewed at least every three years, so its review was due in September 2026. It is cited by Brightwater, Terrasol and Larkspur Capital. One click asks Lena Ortiz for the reviewed policy; when it lands in the library, every answer that cites it is re-checked at once.
“Cited in 3 open questionnaires. When it arrives, each one is re-checked at once.”
For Brightwater's 24 questions the owner router sent 7 sections to 6 owners. Priya has 2 Climate & energy answers ready to approve; Omar Haddad has water with 1 gap; Lena Ortiz has supply chain with 1 flag. Reminders go out after 24 hours, then the ESG lead is copied.
“Owners approve only their own sections. Public environmental claims also need Grace Liu (Legal). Only approved answers can be exported.”
Eight questionnaires submitted this year at 11.6 workdays each on average, against the survey baseline of 18; 54% of answers reused. Before anything reached a rating provider or customer, the checks caught 31 issues: 9 figures that disagreed with the facts register, 4 unsupported environmental claims, 7 documents past their review date and 11 gaps routed to owners.
The same questionnaires, seen by the five people who answer them — what their season looked like, and what it looks like now.
Six agents read each questionnaire, fill it from the library, check every figure and claim and route the rest; section owners approve every answer.
Reads a questionnaire in Excel, Word, PDF or a portal export and finds every question, section, answer format, length limit and required attachment, keeping the original cell or paragraph reference.
Maps each question to the topic it really asks, even when the wording differs, and to the approved answer and library passages that answer it — preferring the newest assured source.
Reuses the approved answer when it fits; otherwise drafts from evidence in the requester's format and length, with a numbered citation for every statement.
Compares every figure with the facts register and every environmental claim with the claim rules, flags documents past their review date, and proposes the fix with both sources side by side.
Sends each section to its owner with the draft, the sources and a due date two working days before the requester's, asks for missing documents and sends reminders.
Writes approved answers back into the requester's own file or portal format, attaches the evidence they asked for and records who approved each answer.
Approve every answer. The agents propose; a named person decides.
Ask in plain words about any request, answer, figure or owner — or say what to change. Every answer points to its source.
Where do we still use the old 61,400 emissions figure?
Three places: Brightwater B.1 (the draft reused the March answer), Clearwater DP-02 (the rater still holds 61,400) and the Halden Health answer submitted on March 20. The assured value is 58,920 tCO2e, restated after the sale of Plant 6. Update everywhere puts the assured figure into both open questionnaires and drafts Halden's correction note for Sam Patel.
Block “carbon neutral” wording in every answer
Done. New rule: any answer saying “carbon neutral”, “climate neutral” or “offset” needs current evidence and Legal's second approval, since the offset programme ended on January 1, 2026. Open drafts that use it are flagged with the approved wording beside them, and the Ostara 2024 answers are marked “do not reuse”. The rule is live in Settings and in the audit trail.
What is our Scope 3 footprint?
Scope 3 was 1.42 million tCO2e in 2025; purchased goods and services are 71%. We screen all 15 categories and report the 12 that are relevant. Suppliers covering 48% of purchased-goods emissions have science-based targets, against 70% by 2028.
Which sites are in water-stressed areas?
Two: Plant 5 Sevilla and Plant 7 Pune, both with site water plans. Total withdrawal was 2.1 million m³ in 2025 and 1.52 million m³ so far in 2026. All four active-ingredient sites apply discharge limits; 2025 effluent results for Plant 4 — Cork are the open gap on Brightwater C.2.
13 screens from the working solution, on its sample data. Pick one to see it large.
Who approves each section and the rules every answer must pass are settings, recorded on the audit trail.
| Setting | Default | Choose from |
|---|---|---|
| Climate & energy section owner | Priya Raman | Priya Raman · Omar Haddad · Lena Ortiz · Dana Okafor · Jonas Weber · Dr. Maya Chen |
| People, safety & access | Jonas Weber; access to medicines answers go to Dr. Maya Chen | Any section owner, per section |
| Figures must match the facts register | On — assured values win | On · Off |
| Block unsupported environmental claims | carbon neutral · climate neutral · eco-friendly · green product · zero impact | Add words; each needs current evidence and Legal's second approval |
| Reuse only answers approved in the last | 12 months | On · Off — older answers go back to their owner |
| Documents must be inside their review date | On | On · Off — a policy past review is flagged and its owner asked |
| Owners approve within | 2 working days | Reminder after 24 hours, then the ESG lead is copied |
| Your own rule | Describe a rule in plain words | Applied to new and open drafts |
Every figure is labelled: a target the solution is built to, an estimate, a typical published result, or a proven one.
“demo” = seen in the working solution, on its sample data · “typical” = published figures (The Conference Board survey of ESG reporting managers, 2018 — about 18 workdays per request, and more than 10 requests a year for most companies) · “estimated” = our estimate of where that baseline's days go. Rules on claims based on offsetting: Directive (EU) 2024/825 and the UK CMA Green Claims Code. People, companies, sites and figures named on this page are fictional — characters and sample data in the working solution.
Software that answers the sustainability questionnaires a company receives — from ESG rating providers, disclosure platforms, customers, tenders and investors — from one approved source. ESG Questionnaire Response keeps one approved answer per topic in an evidence library, fills each new questionnaire from it with a citation for every statement, and sends the rest to the people who own the answers.
Every figure in a draft is checked against the facts register, where assured values win. When a figure is restated, the topic's owner can update it everywhere: every open questionnaire that asks it takes the new value, and a correction note is drafted for any questionnaire already submitted.
Yes. The questionnaire reader handles Excel, Word, PDF and portal exports and keeps the requester's wording and numbering. The export composer writes approved answers back into the requester's own file or portal format, with the evidence they asked for and the approval log.
Claim wording such as “carbon neutral” or “climate neutral” is on a blocked list. The consistency and claim checker flags it in any draft, explains why the current evidence does not support it and offers approved wording; keeping the claim needs Legal's second approval.
The library marks it as review due, and every answer that attaches or cites it is flagged. The policy owner is asked for the current version, and when it lands every answer that cites it is re-checked at once.
Yes. Agents read, match, draft and check; they never approve. Each section owner approves the answers in their section, public environmental claims need the claims counsel's second approval, the ESG reporting lead signs off, and only approved answers can be exported.
They are marked as gaps or new questions, never invented. The owner router sends each one to its owner with the question, what is missing and a due date two working days before the requester's, and the answer joins the library once approved — so the next questionnaire that asks it is filled automatically.
The Agentic Solution Engine builds and deploys it from your requirements and documents — your sustainability report, policies, data extracts and past questionnaire responses — and it goes live once every quality gate has passed. We will walk you through it on one of your own past questionnaires first.
Every relevant bill and rule read overnight and briefed with sources.
1-min demoEarnings Call PreparationLikely analyst questions answered from approved disclosures, cited.
1-min demoPartnering Opportunity TriageEvery inbound pitch scored and answered within days, earlier decisions brought back.
1-min demoWe'll run ESG Questionnaire Response on one of your own past questionnaires and your published report.