AI-native Tools · FactTrace AI
Every claim in a sustainability report, verified against evidence
Upload a report and AI grades every claim against its disclosed evidence. Greenwashing, data gaps, and weakly supported statements surface in about 15 minutes.
Solution intro
FactTrace AI is an AI engine that grades the disclosure integrity of sustainability reports.
It measures how well each claim is backed by disclosed evidence, not a company's underlying ESG performance. Unlike legacy ratings that blend reputation and estimates, FactTrace credits a claim only when evidence supports it. Every verdict traces back to the source page.
MEASURED
VERIFIED
GAP
Common problems in ESG due diligence
“One report takes days to review”
Manual analyst review cannot keep pace with an entire portfolio.
“Legacy ESG ratings are a black box”
Measured data, reputation, and estimates blend together, so the basis of a score is unclear.
“Unverified claims reach our risk models”
Greenwashing passes through unfiltered into investment decisions and compliance reporting.
“Nothing can be fixed after publication”
Issuers need a way to screen weak statements and data gaps before disclosure, not after.
Key features
From upload to verdict in about 15 minutes
Upload, analysis, and improvement requests, in the order you actually work.
Days of review, cut to about 15 minutes
Upload a report and multi-layer AI processes the unstructured document. Gap analysis, missing data, improvement feedback, and regulatory alignment arrive in about 15 minutes.
Evidence grading at the claim level
Each claim is scored on six signals. Measurement, verification, and concrete plans strengthen the evidence, while flowery language, hedged wording, and vague phrasing lower the score. Missing evidence is flagged sentence by sentence.
Where narrative and data disagree
A patented iterative-review LLM finds gaps between what the narrative claims and what the data shows. Every qualitative claim is checked against the quantitative evidence disclosed in the report. Repeated review passes suppress hallucination, so results can be reproduced and used in audits.
From analysis to improvement requests in one flow
View portfolio averages, regional and industry breakdowns, and benchmark comparisons on one screen. Built-in messaging lets you send improvement requests to the analyzed company, citing the findings.
Outcomes
The cost structure of due diligence, changed
~
15
min
analysis time per report
17,000
+
sustainability reports processed
20M
+
ESG benchmark data points
993
standardized metric framework
v2.0
Every verdict traces to the source page, defensible to investment committees and regulators
Single-tenant cloud or on-premise (air-gap) deployment
Customer data is never used to train shared models
Trust
Why trust i-ESG
Certifications and patents
13+ ESG-related patents
ISO 9001/27001
Certified B-Corp
UNGC member
Born as a Fortune 500 in-house venture
Security
AES-256 at rest
TLS 1.3 in transit
Customer data never trains shared models
17,000+
reports analyzed
20M+
benchmark data points
993-metric
standardization framework
Try it on a report from your own portfolio and see the results for yourself.
One report PDF is enough to start.
FAQ
Questions we hear most often during evaluation
How is this different from ESG ratings?
FactTrace measures the disclosure integrity of a report, not ESG performance. It credits only claims that disclosed evidence supports, without reputation or estimates.
Can we trust the analysis?
A patented iterative-review LLM suppresses hallucination, and every verdict cites its source for direct verification.
How is data security handled?
We support single-tenant cloud or on-premise (air-gap) deployment. Customer data is never used to train shared models.
Which report languages are supported?
English is supported by default. Additional languages are available on request.