goAML filings the FIU accepts. The first time.
Thiqara builds every STR, DPMSR, and REAR with a serializer locked to the UAE FIU's own goAML schema and validated against it on every release, so schema errors surface in the product, not at the FIU. Onboarding, screening, monitoring, and reporting, in one UAE-resident platform.
Why it matters
The examiner doesn't grade on effort.
Cabinet Resolution 71/2024 sets out 41 categories of AML fines for DNFBPs, from AED 50,000 to AED 1,000,000 per violation, and supervisors can now fine up to AED 5,000,000 per violation under Federal Decree-Law 10/2025. Failing to report an STR, deliberately or through gross negligence, is a criminal offence. A rejected goAML file, a screening gap, an audit trail with holes: each one is a finding waiting to be written up. Compliance in the UAE isn't a policy document. It's whether, on the day of inspection, every decision on your book has a name, a timestamp, and its evidence attached.
AED 50,000 to 1,000,000
per administrative violation (Cabinet Resolution 71/2024)
AED 100,000 to 1,000,000 + imprisonment
failure to report an STR (Art. 28, Federal Decree Law 10/2025)
Same day
TFS freeze obligation on a UN / UAE Local List match
Regulatory reporting
Rejections happen on your screen. Not at the FIU.
Thiqara ships the UAE FIU's published goAML schema (XSD V2.00, goAML 5.0.2) inside the product. A schema-locked serializer generates every filing, and its output is validated against the FIU's own schema every time the platform changes. There is no freeform drafting to get wrong: close a case with an STR, confirm a freeze, or log a qualifying deal, and the matching report appears in your queue with the trigger and evidence already attached.
<report> <report_code>STR</report_code> <submission_date>2026-07-11</submission_date> <transaction> <amount_local>49400</amount_local> </transaction> </report>
Schema-locked XML, validated on every release
A schema-locked serializer generates the XML, and its output is re-checked against the FIU's XSD every time the platform changes.
The queue remembers so you don't have to
Close a case with an STR, confirm a freeze, or log a qualifying deal, and the matching report appears in your queue with the trigger and evidence attached.
From STR to REAR, one workflow
Suspicious-activity, freeze, threshold, high-risk-country, and name-match filings all move through the same queue and the same submission steps.
STR SAR FFR DPMSR REAR HRC / HRCA PNMRA / CNMRA AIF / AIFT
How it works
From transaction to submitted filing, without leaving the platform.
It watches.
Six monitoring scenarios check every transaction the moment it lands, with no overnight batch. Screening runs on customers, UBOs, and counterparties against live sanctions data, and fails closed if the provider is down.
It builds the case.
An alert opens a case with the evidence pre-attached. Analysts disposition, the MLRO decides; nobody clears their own alert, and the trail shows who did what.
It files.
The report is drafted from the legal trigger through a schema-locked serializer and joins your pickup queue ready to submit. STR, DPMSR, REAR, FFR: same queue, same steps.
Detection & risk
Catch it the moment it lands. Prove why you scored it that way.
Six tunable scenarios (single and linked thresholds, structuring, rapid movement, high-risk countries, dormant reactivation) check every transaction in real time. The 12-factor risk model shows its working, so when the examiner asks why a customer is rated medium, the score breaks down into the exact weights you set. Two floors are deliberately un-tunable as Thiqara policy: a confirmed sanctions match forces high, and PEP exposure floors at medium, built on the enhanced due diligence Article 16 of Cabinet Resolution 134/2025 requires for PEPs.
Screening & sanctions
Fail-closed screening with a same-day freeze path.
Customers, UBOs, and counterparties are screened against live sanctions data. If the provider is unreachable, the platform refuses to pretend there were no hits. Ongoing re-screening checks your whole book against current lists, so a new designation doesn't sit waiting for your next review cycle.
Fail-closed by design
If the screening provider goes down, you get an explicit error. The platform never records a silent “no hits.”
Adverse media, scored
News screening runs automatically when a PEP hit is confirmed, and every article is scored against a fixed AML risk vocabulary so the relevant ones surface first.
Two-tier review ladder
Analysts disposition hits; the MLRO resolves them. Nobody clears their own alert, and the trail shows who did what.
Governance & evidence
Your defence file writes itself.
Every compliance-relevant action is appended to a hash-chained log; edit anything after the fact and the chain visibly breaks. Material decisions (onboarding, EDD approvals, closures, exits) carry a frozen snapshot of the KYC, UBO, and document state exactly as it stood at the moment of decision. Four-eyes approval is enforced in code, not on a checklist. When the examiner asks who decided this, when, and on what evidence, you click once and hand them the chain.
Tamper-evident audit trail
Every compliance-relevant action is appended to a hash-chained log, and any edit after the fact visibly breaks the chain.
Frozen decision snapshots
Material judgements such as onboarding, EDD approvals, case closures, and exits carry the KYC, UBO, and document state exactly as it stood when the decision was made.
Four-eyes, enforced in code
The approver and the senior-management acknowledger of an EDD or AML exit must be different people, and the platform enforces it rather than leaving it to a checklist.
Tipping-off protection, structurally
STR data lives in a restricted schema that ordinary queries can't reach, and report notifications go only to the MLRO tier.
Language
English and Arabic, from day one.
Every screen, tutorial, and notification ships in both languages with full right-to-left layout. Arabic-speaking analysts get the same product as everyone else.
Suspicious Transaction Report
Drafted from case #C-2041 · pending MLRO review · valid against goAML schema 5.0.2
تقرير معاملة مشبوهة
تمت صياغته من القضية C-2041 · بانتظار مراجعة مسؤول الإبلاغ · مطابق لمخطط goAML 5.0.2
Who it's for
Built for the desks that answer to a UAE supervisor.
If you file to the UAE FIU through goAML and answer to a supervisor, this was built for your desk.
Dealers in precious metals & stones
DPMSR threshold filings, from the till
Every cash deal at AED 55,000 crosses the DPMSR threshold. Thiqara surfaces the draft straight from the till record, with the amount, counterparty, and screening result already in place, so a busy Saturday doesn't become a missed filing.
Real estate brokers & developers
REAR without the spreadsheet
Buyer and UBO screening runs before the deal is on your book, and every reportable transaction carries its evidence with it: no spreadsheet, no reconstruction at month-end.
Corporate services, exchange & fintech
An MLRO desk that scales
Roles for the whole function (analyst, deputy MLRO, auditor), with leave-cover delegation and an audit trail the owner can actually read.
Questions
What compliance officers ask first.
How fast can we be live?
One day. If your business's specific complexity calls for a custom implementation, up to ten days.
Does this actually generate valid goAML XML?
Yes. The platform ships the UAE FIU's published XSD (bundle V2.00, goAML schema 5.0.2), generates every filing through a schema-locked serializer, and validates the output against the XSD in our build pipeline on every change. Eleven goAML report codes are supported: STR, SAR, PSTR, DPMSR, REAR, HRC, HRCA, PNMRA, CNMRA, AIF and AIFT, and funds-freeze reports are drafted from TFS matches and filed under the FIU's TFS indicators.
Where does our data live?
Your data is UAE-resident, isolated per organization with row-level security in Postgres. STR and report data additionally sits in a restricted schema that ordinary queries can't reach, supporting your tipping-off obligations.
Can we tune the detection rules, or are we stuck with defaults?
Every scenario is configurable: thresholds, time windows, the risk-model weights, and the rating boundaries. Two floors are deliberately not tunable: confirmed sanctions matches force a high rating, and PEP exposure floors it at medium.
How does it handle TFS obligations?
Continuous re-screening checks your customer book against current lists. A match on a mandatory list (UN Consolidated or UAE Local Terrorist List) triggers an automatic freeze, opens an investigation case, and drafts the funds-freeze report for submission to the Executive Office (EOCN) the same day.
We're a small team. Do we need a big compliance department to run this?
No. One MLRO and one analyst is a complete setup: the pickup queue tells you what needs filing, the dashboard rail tells you what needs attention first, and bilingual in-app tutorials cover every workflow. Deputy delegation covers leave without handing over the whole function.
What happens if the screening provider goes down?
Screening fails closed. You get an explicit error and the action blocks; the platform will never silently record “no hits” because a provider was unreachable.