The Problem

Manual FOI Processing Cannot Meet Statutory Deadlines at Volume

Most agencies use keyword search across disconnected repositories — missing documents in non-indexed folders, scanned files, and archive systems. Manual review queues grow faster than team capacity. Compliance risk accumulates with every missed deadline.

Incomplete Repository Search
Keyword search misses synonyms, redacted terms, and documents in archive systems. No cross-repository search. Completeness cannot be guaranteed.
Statutory Deadline Breaches
Manual processing cannot scale with request volume. Deadline breaches trigger compliance notices, public reporting obligations, and potential legal exposure.
Fragmented Document Repositories
SharePoint, iManage, Trim, and TechONE each require separate searches. No cross-system query capability. Relevant documents are routinely missed.
Request Volume Outpacing Capacity
FOI request volume has increased 300% in five years. Manual teams cannot scale linearly. Backlog compounds and deadline risk grows each quarter.
AI4FOI vs Manual Process
Search Time per Request
Manual: 2–5 business days
AI4FOI: 3–8 minutes
Repository Coverage
Manual: ~60% (keyword only)
AI4FOI: 100% (semantic + OCR)
Response Drafting
Manual: 2–4 hours per request
AI4FOI: Auto-generated draft
Legacy & Scanned Files
Manual: Excluded or manual retype
AI4FOI: OCR-indexed, fully searchable
AI4FOI's EnigmaAI engine applies LLM-based semantic search, extractive and abstractive summarisation, and entity recognition — trained specifically on FOI domain data.
EnigmaAI Engine

EnigmaAI Technical Architecture Domain-Trained LLM Engine for FOI & Records Management

Our AI engine, meticulously designed and engineered in Australia, is automating the document discovery, classification, and response generation steps that consume the most analyst time — leveraging LLMs like GPT and BERT fine-tuned specifically for FOI domain knowledge.

Developed in Australia
Built and maintained by an Australian engineering team — tested against Australian FOI Act and state equivalents. Designed for Australian government data sovereignty requirements.
Zero Infrastructure Deployment
SaaS delivery. No server provisioning, no software installation, no VPN required. Connectors to SharePoint and Trim are configured via admin portal — go live in days.
Tier IV Hosting — Perth, Australia
Hosted in a Tier IV data centre in Perth, Western Australia. 99.99% uptime SLA. Australian data residency — data does not leave Australian jurisdiction.
Domain-Fine-Tuned LLM
GPT-4 and BERT fine-tuned on FOI Act documents, tribunal decisions, and exemption case law. Recognises section references, redaction markers, and third-party consultation requirements.

Unlock the Benefits

  • Efficiency: Streamline Records Management and FOI processes with the power of AI, saving time and resources.
  • Accuracy: Precise and accurate insights — minimising errors and maximising the completeness of FOI responses.
  • Accessibility: Seamless access to information with a user-friendly platform designed to meet specific organisational needs.
  • Versatility: Communicates with diverse document repositories — adaptable to different organisational setups.
Core Capabilities

8 Technical Modules — End-to-End FOI Automation

By integrating these modules, AI4FOI EnigmaAI engine effectively streamlines the process of responding to FOI requests — offering efficiency, accuracy, and convenience.

01
Document Parsing

Parse and extract text from emails, PDFs, Word documents, and spreadsheets. OCR handles scanned documents — no document is missed regardless of format.

02
Natural Language Processing

Entity recognition, sentiment analysis, and topic modelling to understand the content and context of documents — beyond simple keyword matching.

03
Information Retrieval

Searches SharePoint, iManage, Trim, and TechONE ECM. Documents indexed, queried, and ranked by relevance — comprehensive discovery across all sources.

04
LLM Fine-tuning

GPT and BERT fine-tuned on domain-specific FOI data — delivering accuracy and contextual understanding that generic AI models cannot match.

05
Summarisation

Extractive summarisation selects key sentences; abstractive summarisation generates new summaries — providing decision-makers with concise, accurate overviews.

06
Workflow Automation

Orchestrates the entire FOI response process — document search, summarisation, response generation, communication, and approval workflows within the organisation.

07
Repository Integration

Native connectors for SharePoint, iManage, Trim, and TechONE ECM — integratedly accessing documents via the APIs and protocols each system supports.

08
User Interface

A user-friendly interface for government organisations and local councils to submit requests, track status, and access AI-generated responses — all in one place.

Supported Repositories

Works With the Systems You Already Use

AI4FOI integrates with diverse document management systems used across government and public sector — no matter where your documents live.

Microsoft SharePoint
iManage
Trim / HP Records Manager
TechONE ECM
Email Archives
File Servers & Network Drives
+ Additional via API Connectors
How It Works

From Request to Response in 4 Steps

1

Submit

FOI request received via the user interface, email, or web portal. EnigmaAI parses and understands the request using NLP.

2

Search

AI automatically searches all connected repositories, indexes documents, and ranks results by relevance to the request.

3

Summarise

LLM-powered summarisation condenses discovered documents into concise, accurate summaries — both extractive and abstractive.

4

Respond

Workflow engine orchestrates review, approval, and response generation. The final FOI response is produced accurately and on time.

Deployment Results

Deployment Metrics Across Government Environments

AI4FOI is not a generic document Q&A tool — it handles exemption classification, third-party consultation triggers, and audit trail generation as first-class features, not afterthoughts.

Minutes
Search Time per Request
Semantic search across SharePoint, iManage, Trim, and TechONE simultaneously. Results ranked by relevance — not just keyword frequency.
100%
Repository Coverage
OCR pipeline processes tiff, jpg, pdf, and legacy paper scans. Every document in every connected repository is indexed and searchable.
Zero
On-Premise Infrastructure
SaaS delivery from Tier IV data centre. No server, no VPN, no client software. Connector configuration completed via web portal — live within days of onboarding.
NLP
Semantic Query Depth
Named entity recognition, topic modelling, and semantic embedding search. Understands request intent — returns relevant documents even when exact terms differ.

Request a Technical Walkthrough on Your FOI Stack

Tell us which repositories you use — SharePoint, Trim, iManage, or TechONE. We'll walk through connector setup, search accuracy on your document types, and the compliance audit trail before you commit.

Request a Demo Contact Our Public Sector Team

enquiries@datacynical.com · 1 Business Day Response · Singapore & Global