Data Collection & Workflow Systems
Forms, intake systems, workflow automation, structured databases and the route from raw submissions to usable records.
View hub section
Resources for evidence workflows, AI retrieval, data collection and reporting systems
Practical guides, case breakdowns and workflow resources for teams dealing with messy source material, weak evidence trails, slow reporting, AI retrieval problems and public-facing outputs.
Browse by topic, search the full library, or use the content hubs to follow a problem from intake through to reporting and decision support.
Content hubs
AI Retrieval & Knowledge Bases
Preparing documents for AI retrieval, building knowledge bases, checking outputs and keeping source material traceable.
View hub sectionSource Traceability & Evidence Workflows
Source registers, quote banks, evidence tables, review workflows and keeping claims linked to the material behind them.
View hub sectionPublic Submissions & Qualitative Synthesis
Public consultation analysis, stakeholder submissions, interviews, case studies, coding and report-ready findings.
View hub sectionReporting, M&E & Decision Workflows
Reporting bottlenecks, M&E evidence, findings-to-recommendations workflows and decision-ready outputs.
View hub sectionProcess Failure Case Studies
Real-world breakdowns of data misuse, weak QA, document governance failures, broken evidence trails and process failures.
View hub sectionAll articles
Browse the full resource library
Search all articles or filter by the problem area you are working through.
Filter by hub
Fashion for Relief: When a Charity Cannot Clearly Show Where the Money Went
Fashion for Relief raised almost £4.8 million but spent only 8.5% of its expenditure on charitable grants. A UK regulator found weak records, unauthorised trus…
Feeding Our Future Fraud: When Fake Attendance Records Became Proof of Delivery
Feeding Our Future used fake attendance lists, meal counts and invoices to support claims for meals that were never served. More than $240 million was fraudule…
Modest Needs Foundation Fraud: When Charity Oversight Exists Only on Paper
Modest Needs founder Keith Taylor admitted stealing more than $2.5 million in donations intended for low-income families. He also admitted creating the appeara…
South Africa’s Diesel Price Error: What 0.93 Instead of 93.00 Shows About Human Review
A diesel levy reduction of 93.00 cents per litre was captured as 0.93 cents during South Africa’s May 2026 fuel-price calculation. The mistake was human. The w…
Vietnam’s Environmental Monitoring Data Manipulation: When Automated Evidence Cannot Be Trusted
Vietnamese investigators found signs of interference or data alteration at 168 of 306 environmental-monitoring stations inspected. The readings were allegedly…
Cambridge English IELTS marking error: what the £875,000 Ofqual fine shows about weak data workflows
What the Cambridge English IELTS marking error and £875,000 Ofqual fine show about weak data workflows, traceability, monitoring and human review.
Cambridge OCR physics exam errors: what the £270,000 Ofqual fine shows about weak QA workflows
What the Cambridge OCR physics exam errors and £270,000 Ofqual fine show about weak QA workflows, source data checks, mark schemes and review control.
ManageMyHealth data breach: what health document leaks show about weak data governance
What the ManageMyHealth data breach shows about document governance, sensitive records, access control, retention rules, audit trails and AI-ready source contr…
Pearson fined £2m by Ofqual: what repeated exam failures show about weak process control
What Pearson’s Ofqual fine shows about repeated assessment failures, weak process control, risk signals, escalation, monitoring and traceability.
The Post Office data breach shows why document publishing needs a QA workflow
What the Post Office data breach shows about document publishing QA, redaction, version control, sensitivity checks and controlled public release workflows.
Robodebt Failed Because Income Averages Were Treated as Proof of Debt
Robodebt shows what happens when the wrong data points are used to make serious decisions, human review is weakened, and a system turns income estimates into d…
South Africa’s AI Policy Failed Because the Evidence Trail Broke
South Africa’s withdrawn AI policy shows why evidence-heavy public work needs source traceability, citation checking, human review, and controlled AI workflows.
What Metadata Fields Matter for AI Retrieval?
Metadata helps AI retrieval systems find the right source material, filter weak results, and trace answers back to approved documents.
How to Build a Source-Linked Evidence Table for a Report
Learn how a source-linked evidence table connects sources, evidence excerpts, claims, findings, recommendations and report sections.
How to Build a Quote Bank for Qualitative Reporting
A practical guide to building a quote bank that links interview, case study, and fieldwork quotes to themes, findings, source IDs, and report sections.
How to Build a Source Register for an Evidence-Heavy Report
A practical guide to building a source register that tracks source IDs, file links, review status, themes, sensitivity, and report use for evidence-heavy repor…
How to Prepare a Findings-to-Recommendations Matrix
A practical guide to building a findings-to-recommendations matrix that links evidence, findings, implications, recommendations, owners, priorities, and review…
How to QA an AI Knowledge Base Before a Team Starts Using It
A practical QA checklist for testing an AI knowledge base before launch, covering source quality, retrieval, citations, sensitive data, user rules, and human r…
Community-Led Monitoring in South Africa: Turning Public Feedback into Evidence and Action
Use this when public feedback needs to become coded, reviewable evidence rather than disappearing into notes, inboxes, and summaries.
Monitoring and Evaluation Reporting Workflows for South African NPOs
Start here when M&E data, field notes, case studies, and programme records need a clearer route into funder-ready reports.
SMART Indicators for South African NPOs: From Vague Outcomes to Report-Ready Evidence
Read this when outcomes are too vague to collect consistently or use confidently in reports.
Theory of Change for South African NPOs: From Programme Logic to Report-Ready Evidence
Use this when a Theory of Change needs to become a working evidence system, not a diagram filed away after the proposal.
How to Turn Interviews and Case Studies Into Report-Ready Findings
Learn how to turn interviews, case studies, notes, and source documents into clear findings, evidence matrices, recommendations, and report-ready report sectio…
Why Your Website Form Is Not Enough if the Lead Process Behind It Is Broken
A website form can capture enquiries, but it will not fix a broken lead process. Learn how structured lead records, routing, CRM automation, follow-up tasks, a…
What a Public Consultation Response Matrix Should Include
Learn what a public consultation response matrix should include, how to structure responses, and how to connect public feedback to evidence, decisions, actions…
Why AI Gives Weak Answers When Source Material Is Messy
AI tools often give weak answers because the source material is outdated, duplicated, vague, or poorly structured. Learn how to prepare cleaner AI-ready knowle…
How to Stop Losing Source Traceability in Evidence-Heavy Reports
A practical source traceability workflow for primary contractors, policy teams, and donor-funded research teams working across interviews, submissions, case st…
7 Best Independent Evidence Synthesis Consultants for Policy, Consultation, and Donor Reporting
Compare 7 independent evidence synthesis consultants for policy, consultation, donor reporting, qualitative synthesis, and systematic review work.
Best Independent Database Architect Consultants
A practical buyer's guide to 7 independent database architect consultants, including who each one is best for and how to choose the right fit.
Decision-Ready Insight: Turning Raw Information into Decision-Ready Work
Learn how to turn raw data, documents, and reports into decision-ready insight with a clear evidence workflow. See the guide and case proof.
How to Prepare Documents for AI Retrieval Without Losing Structure or Traceability
Read this first when the retrieval problem sits in PDFs, spreadsheets, OCR, parsing, metadata, or version mess.
Reporting Workflows: From Evidence to Recommendations
Learn how strong report writing workflows move from evidence planning to synthesis, findings, conclusions, recommendations, and human-reviewed AI support.
How to Build an AI-Ready Knowledge Environment for Internal Retrieval
Use this when the bigger issue is search design, access control, metadata logic, and retrieval architecture across an internal corpus.
How to Build Evidence Workflows for Reporting and Accountability
Start here when the real bottleneck sits between intake, synthesis, review, and final reporting output.
How to Synthesise Stakeholder Submissions Without Losing Source Traceability
Synthesise stakeholder submissions with source IDs, coding, framework matrices, and QA for traceable, defensible reporting.
The Real Cost of Messy Evidence Workflows
Read this when the same reporting cycle keeps turning into cleanup, rework, and late-stage review pain.
Content hubs
Follow the workflow by topic
The blog is organised into resource hubs so you can follow a problem from the first intake point through to evidence structure, AI retrieval, synthesis, reporting and real-world failure examples.
Useful calculators
Estimate time, risk or reporting pressure
Use these tools to estimate where time, risk or reporting pressure may be sitting in your current workflow.
Services
View all servicesNeed help turning the ideas into a working system?
The resources above explain the problems. The services help teams fix them in practice by improving how information is collected, structured, reviewed, reported and used.
Data Collection & Intake Systems
For teams that need better forms, intake workflows, structured records, source IDs, review fields and handover-ready data from the start.
View serviceTraceable Evidence Workflow Support
For teams that already have source material and need to turn it into structured evidence, coded themes, findings, recommendations and report-ready outputs.
View serviceData Use, Reporting & Communication Systems
For teams that need to turn structured information into reports, dashboards, tools, microsites, briefs or public-facing outputs.
View serviceFAQ
Questions about the resource library
Use these answers to choose the right hub, article, calculator or service path.
What is this blog about?
This blog covers practical ways to collect, structure, analyse, review and use information. The focus is on evidence workflows, AI retrieval, source traceability, qualitative synthesis, reporting systems and real-world process failures.
Do you have resources on AI and data?
Yes. The AI Retrieval & Knowledge Bases hub covers AI-ready source material, metadata, document retrieval, weak AI answers, knowledge base QA and source-grounded outputs.
Do you have resources on report writing?
Yes. The Reporting, M&E & Decision Workflows hub covers reporting bottlenecks, findings-to-recommendations matrices, M&E evidence, theory of change, SMART indicators and decision-ready insight.
Do you have resources on source traceability?
Yes. The Source Traceability & Evidence Workflows hub covers source registers, evidence tables, quote banks, review status, claim tracking and the route from source material to report-ready findings.
Do you have resources on public submissions and qualitative data?
Yes. The Public Submissions & Qualitative Synthesis hub covers stakeholder submissions, public consultation response matrices, interviews, case studies, open-text comments, coding and synthesis.
Do you have examples of real-world data or process failures?
Yes. The Process Failure Case Studies hub breaks down public examples where weak data workflows, document governance, QA, evidence trails, AI use or publishing processes created avoidable risk.
How should I choose where to start?
Start with the part of the workflow causing the most friction. If information comes in badly, start with data collection. If sources are hard to trace, start with evidence workflows. If AI answers are weak, start with AI retrieval. If reports are slow, start with reporting workflows.
Let's talk
Need a clearer route from messy information to usable outputs?
If your team is dealing with scattered source material, weak evidence trails, slow reporting, AI retrieval problems or repeated manual review, I can help design the workflow behind the output.