Industries

Built for the fields that show up in your work.

LegalFinanceHealthcareHR & recruitingInsurancePublic sectorCustomer supportAccounting & taxReal estate

Nine industries, one document each. Here's what Synonym catches.

Banks · fintechs · accounting

Finance

An IBAN is masked because its mod-97 checksum passes, not because it looks like one.

Hover a field to see where it goes. Mask the document in place:
Your documentstays with you
Payment note — incoming transferTransfer received to 1SK31 1200 0000 1987 4263 7541 from company ID 235763469, BIC 3TATRSKBX. Counter-account (CZ) 419-2000145399/0800; card on file 54111 1111 1111 1111. Draft a confirmation reply.
tokens only
What the model readsany model
Payment note — incoming transferTransfer received to 1<IBAN_1> from company ID 2<COMPANY_ID_1>, BIC 3<BIC_1>. Counter-account (CZ) 4<BANK_ACCOUNT_1>; card on file 5<CREDIT_CARD_1>. Draft a confirmation reply.
1IBAN → <IBAN_1> mod-972Company ID / Tax ID / VAT → <COMPANY_ID_1>3BIC / SWIFT → <BIC_1>4Bank account (CZ) → <BANK_ACCOUNT_1> checksum5Credit card → <CREDIT_CARD_1> Luhn
Detector pack · Finance
SignatureIBAN · the account behind every transfer
Checksum-verified
  • IBAN
  • Bank account (CZ)
  • Credit / debit card
  • SEPA Creditor Identifier
  • Company ID
  • Tax ID / VAT ID
  • LEI
  • ISIN
  • EORI number
  • Ukrainian tax ID (RNOKPP)
  • Neighbouring national IDs
Format-verified
  • BIC / SWIFT
  • Card expiry & CVV
  • Payment symbols (VS / KS / SS)
  • Crypto wallet
  • Transaction hash
  • ID card number
  • Ukrainian passport / ID
  • Custom pattern detector
Language model
  • Organisation
live todaylanguage-model detectors rolling out
Clinics · insurers · care coordination

Healthcare

Diagnosis codes and insurance IDs are opt-in. Identity and diagnosis are held back.

Hover a field to see where it goes. Mask the document in place:
Your documentstays with you
Referral — outpatient clinicPatient 1Ján Novák, born 2900101/0007, insurance ID 39001010007, 4Hlavná 12, Košice. Diagnosis 5J45.0, BP 6140/90 mmHg. Turn these notes into a referral letter.
tokens only
What the model readsany model
Referral — outpatient clinicPatient 1<PERSON_1>, born 2<BIRTH_NUMBER_1>, insurance ID 3<INSURANCE_ID_1>, 4<LOCATION_1>. Diagnosis 5<DIAGNOSIS_1>, BP 6<BLOOD_PRESSURE_1>. Turn these notes into a referral letter.
1Patient & next-of-kin names → <PERSON_1>2Birth number → <BIRTH_NUMBER_1> checksum3Insurance holder ID → <INSURANCE_ID_1> opt-in4Home address / location → <LOCATION_1>5Diagnosis codes → <DIAGNOSIS_1> opt-in6Blood pressure → <BLOOD_PRESSURE_1> opt-in
Detector pack · Healthcare
SignatureDiagnosis code · the field that makes a note sensitive
Checksum-verified
  • Birth number
Format-verified
  • EHIC card number
  • Date of birth
  • Age
  • Custom pattern detector
Registry lookup
  • Health insurer code
  • Drug / ATC code
Opt-in
  • Diagnosis code
  • Insured-person ID
  • Lab result with unit
  • Blood type
  • Blood pressure
Language model
  • Person name
  • Location / address
  • Genetic / disability mention
live todaylanguage-model detectors rolling out
Recruiters · HR teams · payroll

HR & recruiting

Identity documents and contact details are held back. The role and the checklist stay readable.

Hover a field to see where it goes. Mask the document in place:
Your documentstays with you
New hire — senior accountantCandidate 1Ján Novák, born 2900101/0007, 3jan.novak@alfa.sk, 4+421 903 123 456, passport 5AB1234567, lives at 6Hlavná 12, Košice. Strong IFRS background. Draft the onboarding checklist.
tokens only
What the model readsany model
New hire — senior accountantCandidate 1<PERSON_1>, born 2<BIRTH_NUMBER_1>, 3<EMAIL_1>, 4<PHONE_1>, passport 5<PASSPORT_1>, lives at 6<LOCATION_1>. Strong IFRS background. Draft the onboarding checklist.
1Candidate names → <PERSON_1>2Birth number → <BIRTH_NUMBER_1> checksum3E-mail addresses → <EMAIL_1>4Phone numbers → <PHONE_1>5Passport number → <PASSPORT_1> check digit6Home address → <LOCATION_1>
Detector pack · HR & recruiting
SignaturePassport & ID card · the documents in every hire
Checksum-verified
  • Passport
  • Birth number
  • Ukrainian tax ID (RNOKPP)
  • Neighbouring national IDs
Format-verified
  • ID card number
  • Driving licence number
  • Date of birth
  • Age
  • Ukrainian passport / ID
  • EHIC card number
  • E-mail address
  • Phone number
  • Profile URL / handle
  • URL with personal path
  • Salary amount
  • Custom pattern detector
Language model
  • Person name
  • Location / address
  • Criminal record reference
  • Genetic / disability mention
  • Union / religion / health mention
  • Gender / nationality / ethnicity mention
  • Political / sexual-orientation mention
live todaylanguage-model detectors rolling out
Insurers · brokers · claims handling

Insurance

Identity, payout account and incident address are masked before a claim is summarised.

Hover a field to see where it goes. Mask the document in place:
Your documentstays with you
Claim — motor policy 77-4412Policyholder 1Ján Novák, birth number 2900101/0007, insured no. 39001010007. Incident at 4Hlavná 12, Košice. Payout to 5SK31 1200 0000 1987 4263 7541. Assess liability and suggest next steps.
tokens only
What the model readsany model
Claim — motor policy 77-4412Policyholder 1<PERSON_1>, birth number 2<BIRTH_NUMBER_1>, insured no. 3<INSURANCE_ID_1>. Incident at 4<LOCATION_1>. Payout to 5<IBAN_1>. Assess liability and suggest next steps.
1Policyholder names → <PERSON_1>2Birth number → <BIRTH_NUMBER_1> checksum3Insured-person ID → <INSURANCE_ID_1> opt-in4Incident address → <LOCATION_1>5Payout IBAN → <IBAN_1> mod-97
Detector pack · Insurance
SignatureVIN · the vehicle behind every claim
Checksum-verified
  • VIN
  • Birth number
  • IBAN
  • Credit / debit card
Format-verified
  • Licence plate
  • Driving licence number
  • ID card number
  • Date of birth
  • Age
  • EHIC card number
  • Phone number
  • Court file number
  • Cadastral parcel / LV
  • GPS coordinates
  • Custom pattern detector
Registry lookup
  • Health insurer code
  • Drug / ATC code
Opt-in
  • Insured-person ID
  • Diagnosis code
Language model
  • Person name
  • Location / address
  • Genetic / disability mention
live todaylanguage-model detectors rolling out
Municipalities · agencies · registries

Public sector

Officials draft replies with an LLM while the citizen stays anonymous to it.

Hover a field to see where it goes. Mask the document in place:
Your documentstays with you
Request — building permit no. 2024/318Applicant 1Ján Novák, birth number 2900101/0007, acting for 3Alfa s.r.o. (company ID 435763469), property at 5Hlavná 12, Košice, contact 6jan.novak@alfa.sk. Prepare the decision notice.
tokens only
What the model readsany model
Request — building permit no. 2024/318Applicant 1<PERSON_1>, birth number 2<BIRTH_NUMBER_1>, acting for 3<ORGANIZATION_1> (company ID 4<COMPANY_ID_1>), property at 5<LOCATION_1>, contact 6<EMAIL_1>. Prepare the decision notice.
1Applicant names → <PERSON_1>2Birth number → <BIRTH_NUMBER_1> checksum3Organizations → <ORGANIZATION_1>4Company ID → <COMPANY_ID_1>5Addresses → <LOCATION_1>6E-mail addresses → <EMAIL_1>
Detector pack · Public sector
SignatureBirth number · the citizen behind every file
Checksum-verified
  • Birth number
  • Passport
  • Company ID
  • Tax ID / VAT ID
  • VIN
  • Ukrainian tax ID (RNOKPP)
  • Neighbouring national IDs
Format-verified
  • ID card number
  • Driving licence number
  • Date of birth
  • Court file number
  • ECLI decision identifier
  • Cadastral parcel / LV
  • GPS coordinates
  • Licence plate
  • E-mail address
  • Ukrainian passport / ID
  • Custom pattern detector
Registry lookup
  • Data box ID (CZ)
Language model
  • Person name
  • Organisation
  • Location / address
  • Criminal record reference
  • Gender / nationality / ethnicity mention
live todaylanguage-model detectors rolling out
Telcos · utilities · e-commerce

Customer support

Agents summarise, translate and draft replies without contact details reaching the model.

Hover a field to see where it goes. Mask the document in place:
Your documentstays with you
Ticket #48213 — refund requestCustomer 1Ján Novák (2jan.novak@alfa.sk, 3+421 903 123 456) asks for a refund to card 44111 1111 1111 1111; last login from 585.237.12.9. Summarise and draft a reply.
tokens only
What the model readsany model
Ticket #48213 — refund requestCustomer 1<PERSON_1> (2<EMAIL_1>, 3<PHONE_1>) asks for a refund to card 4<CREDIT_CARD_1>; last login from 5<IP_1>. Summarise and draft a reply.
1Customer names → <PERSON_1>2E-mail addresses → <EMAIL_1>3Phone numbers → <PHONE_1>4Card on file → <CREDIT_CARD_1> Luhn5IP addresses → <IP_1>
Detector pack · Customer support
SignatureIMEI & SIM ICCID · the device behind every ticket
Checksum-verified
  • IMEI
  • SIM ICCID
  • Credit / debit card
  • IBAN
  • SEPA Creditor Identifier
  • Electricity / gas supply point (EIC)
  • Parcel tracking number
  • GS1 / EAN code
  • VIN
Format-verified
  • MAC address
  • IP address
  • E-mail address
  • Phone number
  • Profile URL / handle
  • URL with personal path
  • Session tokens & secrets
  • Card expiry & CVV
  • Payment symbols (VS / KS / SS)
  • Crypto wallet
  • Custom pattern detector
Language model
  • Person name
live todaylanguage-model detectors rolling out
Accountants · tax advisors · auditors

Accounting & tax

Company ID, VAT number and bank account each have a fixed, validated format.

Hover a field to see where it goes. Mask the document in place:
Your documentstays with you
Query — supplier invoice 2024-0917Supplier 1Alfa s.r.o., company ID 235763469, VAT ID 3SK2020123456, account 4SK31 1200 0000 1987 4263 7541, CZ account 519-2000145399/0800. Is the VAT treatment on this invoice correct?
tokens only
What the model readsany model
Query — supplier invoice 2024-0917Supplier 1<ORGANIZATION_1>, company ID 2<COMPANY_ID_1>, VAT ID 3<VAT_1>, account 4<IBAN_1>, CZ account 5<BANK_ACCOUNT_1>. Is the VAT treatment on this invoice correct?
1Supplier / client firms → <ORGANIZATION_1>2Company ID → <COMPANY_ID_1> check digit3Tax ID / VAT (Tax ID / VAT ID) → <VAT_1> mod 114IBAN → <IBAN_1> mod-975Bank account (CZ) → <BANK_ACCOUNT_1> checksum
Detector pack · Accounting & tax
SignatureTax / VAT ID · the counterparty on every invoice
Checksum-verified
  • Tax ID / VAT ID
  • Company ID
  • IBAN
  • Bank account (CZ)
  • SEPA Creditor Identifier
  • LEI
  • ISIN
  • EORI number
  • GS1 / EAN code
Format-verified
  • BIC / SWIFT
  • Payment symbols (VS / KS / SS)
  • Excise (SEED) number
  • Salary amount
  • Custom pattern detector
Registry lookup
  • Data box ID (CZ)
Language model
  • Organisation
live todaylanguage-model detectors rolling out
Agencies · property managers · notaries

Real estate

Owner, tenant, property and rent account — a lease reviews just as well with tokens in their place.

Hover a field to see where it goes. Mask the document in place:
Your documentstays with you
Lease draft — office spaceLandlord 1Ján Novák, birth number 2900101/0007. Tenant 3Baltrans Logistics, company ID 435763469. Premises: 5Hlavná 12, Košice. Rent payable to 6SK31 1200 0000 1987 4263 7541. Review the termination clause.
tokens only
What the model readsany model
Lease draft — office spaceLandlord 1<PERSON_1>, birth number 2<BIRTH_NUMBER_1>. Tenant 3<ORGANIZATION_1>, company ID 4<COMPANY_ID_1>. Premises: 5<LOCATION_1>. Rent payable to 6<IBAN_1>. Review the termination clause.
1Owner / tenant names → <PERSON_1>2Birth number → <BIRTH_NUMBER_1> checksum3Companies → <ORGANIZATION_1>4Company ID → <COMPANY_ID_1>5Property address → <LOCATION_1>6Rent IBAN → <IBAN_1> mod-97
Detector pack · Real estate
SignatureCadastral parcel / LV · the owner is one lookup away
Checksum-verified
  • Birth number
  • Company ID
  • IBAN
  • Bank account (CZ)
  • Electricity / gas supply point (EIC)
Format-verified
  • Cadastral parcel / LV
  • GPS coordinates
  • ID card number
  • Payment symbols (VS / KS / SS)
  • Custom pattern detector
Language model
  • Person name
  • Organisation
  • Location / address
live todaylanguage-model detectors rolling out

Don't see your industry?

The same detectors work for any workflow with names, account numbers or personal data in it. Try it on your own document, or tell us what you handle.