Prototype / Sample data
Export

Workspace

Settings

What this workspace monitors and where the data comes from.

Monitored brand

One workspace per client

CA

Canara Bank

Public Sector Banking

Active

Data source

Prototype dataset

DatasetCanara Bank · 2,823 captured records
Records in use2,821
Earliest mention29 Jan 2012
Latest mention15 Sep 2026
Engagement figuresNot available in the source data
Live scrapingTo be connected

Platforms monitored

App StoreFacebookInstagramMeta AdsYouTube

Categories tracked

Brand / CampaignComplaintsCustomer ServiceDigital BankingFraud / ScamGeneralJobs / RecruitmentLoansNews / MediaOffers / PromotionsPayments / UPIProduct / Service

Topics detected

33 in this dataset

Account openingBanking app enquiryBranch service complaintBrand campaignBrand mentionBusiness loan enquiryCard servicesDeposits & KYCDigital onboardingExam & interview prepFake loan agentFinancial results & newsFraud alertFund transfer issueGeneral brand mentionGeneral complaintHelpline / contact enquiryHome loan enquiryJob enquiryJob vacancyLoan enquiryMedia coverageMobile banking issueNet banking issuePersonal loan enquiryPhishing / OTP fraudPositive customer experienceProduct enquiryPromotional offerRecruitment updateService enquiryThird-party promotionUPI issue

Alert thresholds

Applied to every incoming mention

High risk score55 and above
Medium risk score28 – 54
Fraud / scam wordingAlways high risk
Negative spike alertOn
Viral negative mention alertOn

Where each field comes from

Every field shown in the product, and how it was produced from the source file

Shown asSource fieldHow it is produced
Postad_descriptionUsed as-is, trimmed for display only.
Original post linksource_urlDirect link to the post on Facebook, Instagram, YouTube, etc.
PlatformplatformUsed as-is.
Accountsm_handler / channel_url / sub_publisherHandle where present; for YouTube the channel handle is read from the channel link, with the channel name alongside it.
Account linkprofile_url / channel_urlOpens the account's page, profile or channel on the platform.
Dateupload_dateThe date the post was captured.
Hashtagsad_descriptionEvery #tag found in the post text.
Categoryad_descriptionKeyword match against 12 ORM categories, General as fallback.
Topicad_description + keyword_termMost specific discussion inside the category.
Sentimentad_descriptionThe whole post is read together: first-hand praise or complaint decides the tone, wording like 'not good' is flipped, and marketing, news or how-to phrasing such as 'best offer' or 'apply now' is not counted as praise. Anything without a clear opinion stays Neutral.
Emotionad_descriptionClosest emotion to the wording and sentiment.
Riskad_descriptionFraud, scam and complaint wording scored 0-100.
Engagement / reachnot availableLikes, comments, shares and views are empty in the source file, so no engagement figure is shown.
AI Summarythe record itselfWritten from the fields above, no external model.