ComplyIQ360 – Frequently Asked Questions

▾

What is a “Standard File”?

A Standard File is the authoritative reference document you upload that defines the criteria, controls or clauses you want to assess against (e.g. a published standard, internal policy manual, supplier requirements, regulatory guideline, ESG / climate disclosure framework). ComplyIQ360 parses this file and indexes its structure so your other documents can be automatically benchmarked clause‑by‑clause.

▾

What kinds of standards can it handle?

Any textual, clause‑based PDF standard or policy in English (multi‑language expansion is on the roadmap). Typical examples include: regulatory (e.g. environmental or financial guidance), governance & risk frameworks, ISO‑style management standards, internal SOPs, supplier codes, cyber / privacy policies, ESG / climate frameworks (e.g. AASB / ISSB style disclosures), and bespoke checklists. If the content is primarily text, structured with sections or numbered items, it will generally work. You have to bring your own standard. The app is currently unable to download or provide an option to purchase the standard text from a third-party website.

▾

What is a “Report File”?

A Report File is any document (draft policy, procedure, supplier document, disclosure draft, etc.) you want assessed against the Standard File. The platform analyses gaps, matches relevant clauses, and generates a structured compliance report summarising alignment, partial matches, omissions and remediation recommendations.

▾

Can I upload multiple files together?

Yes. You may upload: (1) a single Standard File and one or more Report Files in the same job in zipped format.

▾

Why only PDF? Why can’t it handle Word, Excel or images?

PDFs provide a consistent layout and text extraction pipeline, reducing ambiguity in pagination, headings and clause numbering. Other formats (DOCX, XLSX, scanned images) introduce variability (styles, embedded tables, OCR noise) that can degrade clause alignment accuracy. Support for other formats (with conversion + validation) is planned; for now, export to PDF before upload.

▾

Why does it take longer than ChatGPT or MS Copilot?

General chat tools stream quick, shallow responses. ComplyIQ360 orchestrates a multi‑stage pipeline: structural parsing of the standard, semantic chunking, clause–document cross‑matching, scoring, reasoning over gaps, remediation synthesis etc.. This deliberate pipeline trades a few extra minutes for: (a) clause‑level traceability, (b) detailed analysis, and (c) reduced hallucination with complex reasoning. Typical medium jobs complete in a few minutes; very large standards or many report files will take longer.

▾

How reliable are the results?

The system uses large language model reasoning. Each gap or alignment point is grounded in the original text segments stored with metadata (location, clause ID, page reference). This reduces free‑form hallucination.

Strengths: High recall for explicit clauses, and transparent source referencing.

Limitations: Nuanced legal interpretation, jurisdictional subtleties, organisational context, or forward‑looking risk judgements still require human review. Treat the output as a decision support report, not a legal opinion.

We actively monitor error rates; failed jobs auto‑refund credits (see below).

▾

I see the error “user must be logged in to continue”. What should I do?

Your session token likely expired or a transient network/authentication check failed. Usually a simple page reload re‑establishes the session. If that fails, log out and sign back in. Persistent recurrence may indicate a blocked third‑party script or cleared cookies—ensure your browser allows necessary storage.

▾

What happens if I close the browser window while a job is running?

The job continues server‑side. When it finishes, results are stored and accessible from the Job History section of your dashboard. Closing the tab does not cancel processing.

▾

My job finished with an error. Do I get the credit back?

Yes. If a job terminates in an error state (not user‑cancelled), the consumed credit is automatically returned to your balance. You can re‑submit after addressing any file issues (e.g. corrupt PDF, extremely large file, or transient network error).

▾

I am part of an organisation. Which credits are used first?

Organisation credits are consumed before personal (individual) credits when you run a job under that organisation’s context. Once the organisation allocation is exhausted, the platform automatically falls back to your personal balance (if any). The job summary records which pool was debited.

▾

Where is my data stored and for how long?

All files are stored in AWS (Sydney region) with access events logged. When you delete a job, the primary content is removed and only minimal metadata (e.g. anonymised job identifiers, aggregate statistics) may be retained for operational integrity.

Interactions with underlying generative AI services are transient; raw source files (not just extracted snippets) may be transmitted securely for processing, but providers’ standard retention policies (e.g. OpenAI, Gemini, Claude) apply only to ephemeral model operation—we do not permit them to train on your data under standard commercial terms.

Always remove highly sensitive material once your report is finalised if organisational policy requires minimisation.

▾

Best practices for faster & higher quality results?

Clean PDFs: Export directly from the source application (avoid scans / photos).
Split mega‑standards: Extremely large compilations can be segmented logically.
One logical theme per job: Avoid mixing unrelated standards in one run.
Check page order: Ensure no rotated or duplicate pages.
Retry transient errors: If refunded, simply resubmit after a brief wait.

▾

Is the output legal or financial advice?

No. Reports are analytical decision support tools highlighting textual compliance gaps. Always have qualified professionals review and finalise interpretations before regulatory submission or contractual reliance.