Loading...
Ask questions across ESG reports and corporate disclosures. Get sourced answers with exact document citations. No hallucinations. No manual cross-referencing.
Real workflows for analysts, compliance teams, and investors.
Ask questions across 50 company reports at once. Instead of opening each PDF manually, you query your entire document corpus and get cited answers from every relevant source in one go. Scope 3 emissions, governance structures, TCFD alignment: your whole portfolio, one query.
Find TCFD metrics across your entire document set with one query. Check which companies have disclosed ISSB IFRS S2 climate risks, who has board-level governance of sustainability, and where the gaps are. The system maps your queries to the frameworks your institution already uses.
Cross-reference ESG claims in one document against disclosures in another. Verify whether a company's stated emissions reduction targets in its sustainability report match what it reported in its annual filing. Every answer traces back to a specific source so you can check the work yourself.
Four steps from document upload to sourced answer.
Upload PDF reports, Excel files, and regulatory filings directly into your workspace. Supported formats include annual reports, sustainability disclosures, CDP responses, and internal policy documents. Files are stored securely and isolated to your account.
The system chunks, embeds, and indexes your documents with financial-grade precision. Text is segmented into semantically coherent passages, converted into vector representations, and stored in your private index. This typically takes 30 to 120 seconds per document.
Type your question as you would ask a colleague. "What does this company report on Scope 3 emissions?" or "Does this filing disclose TCFD-aligned physical climate risks?" No query syntax, no boolean operators. Just your question.
Every response includes the document name, page number, and the exact text passage the answer was drawn from. You see the generated answer and the source in the same view. Nothing is asserted without a citation.
Choose the right mode for your task. Speed vs. precision is a real trade-off. We give you the controls.
Pure semantic search using dense vector embeddings. The system finds passages that are conceptually similar to your query, even if they use different words. Best for open-ended research questions and exploratory queries where you want broad coverage quickly.
Best for: Initial portfolio screening, exploratory research
Combines dense semantic search with sparse keyword matching (BGE-M3 learned sparse model). This catches both conceptually related passages and exact term matches. Better recall than dense-only, with minimal speed penalty. Good default for most ESG workflows.
Best for: Standard due diligence queries, regulatory alignment checks
Fetches a broad candidate pool, then re-ranks using ColBERT's MaxSim algorithm, which computes late-interaction similarity scores between query tokens and document tokens. The most accurate retrieval mode, designed for critical due diligence where precision matters more than speed.
Best for: Investment committee memos, regulatory submissions, final sign-off documents
The search engine is designed around the frameworks that institutional investors and listed companies actually use. Your queries can reference framework-specific terminology and the system understands the context.
Upload whatever documents your team already works with. The ingestion pipeline handles heterogeneous document sets without requiring a specific format or structure.
Asset managers, compliance officers, private equity analysts, and corporate sustainability teams all use the ESG Search Engine to cut research time and produce sourced, auditable outputs. Start free with 25 searches per month.