Skip to main content
Writing a funder-ready report from raw field submissions can take a programme team weeks — pulling figures from spreadsheets, reconciling inconsistencies, and translating operational data into a coherent narrative. Terriqon’s AI report engine compresses that process to minutes. It reads your validated, sanitised submission data, generates a structured narrative in the appropriate register for your audience, and then puts a human approver in the loop before anything leaves the platform. The result is a report you can trust, produced at a speed your team can actually work with.
AI reports are based on aggregated, sanitised metrics — your beneficiary data and personal information are never sent to the AI.

​
Report Types

Project Report

Summarises submission data, KPIs, and field activity for a single project. Use this for internal progress reviews or project-level accountability reporting.

Ops Report

An operational summary of team performance, submission rates, and data quality for managers. Designed to surface bottlenecks and flag Field Officers who may need support.

Donor Report

A funder-ready narrative covering KPIs, outcomes, and impact metrics. Written in formal language suitable for direct submission to donors, grant bodies, or government partners.

Portfolio Report

A cross-project summary for programme directors overseeing multiple projects. Rolls up KPIs and trends across your entire portfolio into a single consolidated document.

​
How a Report Is Generated

1

Validated submissions are collected

The report engine checks whether enough high-quality submissions are available to produce a meaningful report. Submissions that score below the confidence threshold — due to incomplete fields, outlier values, or other data quality signals — are automatically routed to a human reviewer instead of being passed to the AI. This prevents a low-quality data set from producing a misleading report.
2

The PII stripping pipeline runs

Before any data reaches the AI, Terriqon’s PII stripping pipeline removes all personal and sensitive information. It excludes every field tagged as Personally Identifiable or Restricted, and always strips GPS coordinates regardless of how the field is tagged. Free-text responses, names, phone numbers, device IDs, case IDs, uploaded files, audio recordings, and signatures are also excluded. The pipeline additionally pattern-detects untagged values that match national ID numbers, passport numbers, and bank account numbers, and removes those too. Only anonymised, aggregated metrics proceed to the next step.
3

The AI generates a narrative

The AI model reads the sanitised metrics and generates a structured report draft in the format appropriate for the report type you selected. Section headings, narrative language, and data presentation are all generated automatically.
4

A manager reviews and approves the draft

The draft is held in the platform and cannot be downloaded or shared until a named Organisation Admin or Manager reviews it. Reviewers can read, edit, and annotate every section of the draft before making a decision.
5

The report is approved and made available

Once a manager approves the report, it becomes available to download as a PDF or share via a secure link. The approval — including the reviewer’s name and timestamp — is permanently recorded in the report’s audit trail.

​
The Approval Requirement

No report can be downloaded, shared, or exported without a named Organisation Admin or Manager formally approving it. This is not optional and cannot be bypassed. If a reviewer rejects a draft, they are required to provide a written reason before the rejection is logged. The rejected draft, the reason, and the reviewer’s identity are all stored permanently in the report’s audit trail. A new draft can then be triggered with corrected data or revised parameters. The approval trail gives your organisation a defensible record of human oversight for every report that leaves the platform — important for organisations subject to donor accountability requirements or data governance frameworks.

​
AI Accuracy and Your Editorial Control

Terriqon’s AI is a decision-support tool, not a replacement for programme expertise. It produces structurally sound, contextually appropriate drafts — but it does not have knowledge of your specific programme history, local context, or organisational voice. The platform is honest about data quality: if the underlying submissions are incomplete or inconsistent, the system flags the issue and routes the data to human review rather than generating a confident-sounding report from weak evidence. You have full editorial control over every draft before it is approved. Sections can be rewritten, figures can be corrected, and tone can be adjusted to match your audience.
You can edit any section of an AI draft before approving. Use the editor to add context, correct figures, or adjust tone for your specific audience.