SemanticRisk • Public Company Record
ti.com
A 14-day record of the website’s deterministic interpretation score, aggregate score movement, and claim-level drift. These observations describe this monitoring pipeline and are not a guarantee of downstream model behaviour.
How to read this page: The score is generated from a deterministic weighted evaluation, so fixed score intervals and shared values can reflect the same evaluation outcomes. Score movement describes changes in that aggregate score. The timeline below shows one row per scan observation, while the company explorer calculates material claim drift across stored model comparisons. A single scored scan can produce separate Claude, Gemini, GPT and Grok comparisons. A held observation carries forward the previous valid score when normalized captured content is unchanged.
Why can claims move while the score stays the same? Aggregate scores summarize several evaluation dimensions. Claim-level differences can still occur in attribution, specificity, scope, classification, quantities, references, or wording without changing the total score. Not every extracted difference is material. See how SemanticRisk classifies differences →
Current score ⓘ
57.75 /100
Middle-scoring range
Latest score change ⓘ
No score change
Previous score: 57.75 /100
14-day scan observations with material flag ⓘ
0.0%
0 of 2 displayed scan observations
Each of the 1 newly scored scan observation included at least one material model comparison.
14-day score movement ⓘ
0.000
0 of 1 newly scored comparisons changed the aggregate score
Latest interpretation comparison
What changed, how significant it appears, and the recorded likely driver.
Classification ⓘ
No Change
Likely driver ⓘ
No Material Change
Claims added / removed ⓘ
0 / 0
Stance / confidence changes ⓘ
0 / 0
A changed extraction does not automatically mean the company changed its position. It can reflect source wording, page structure, captured content, or interpretation variation. Review the evidence preview and content-capture status together. See the difference taxonomy →
Crawl and capture health
These signals describe whether enough usable website content was captured to support the interpretation record.
Preflight status ⓘ
Ok
TTFB ⓘ
411 ms
Estimated readable text ⓘ
3,950
Normalized content ⓘ
3,943
14-day score and interpretation record
Bars show the effective score out of 100. A carried-forward marker identifies observations where unchanged normalized content retained the previous valid score. Table rows show score movement separately from claim-level classification.
57.75
57.75
↔ indicates a carried-forward score: normalized captured content was unchanged and no new scored evaluation replaced the previous valid result.
| Scan time | Score /100 ⓘ | Score change ⓘ | Claim classification ⓘ | Severity ⓘ | Status | Evidence |
|---|---|---|---|---|---|---|
| 2026-08-01 10:18 | 57.75 | No score change | No Change | LOW | Ok | No public excerpt |
| 2026-08-01 10:14 | 57.75 | — | Unknown | — | Ok | No public excerpt |
Continue monitoring
14 days shows the current pattern. Longer history helps determine whether it persists.
Paid monitoring adds alerts, longer history, scheduled summaries, competitor comparisons, exports, and validated claim-level evidence with difference categories as classification support expands.