Cognitive scaffolding for qualitative rigour

Your interpretation. Fully traced, methodologically defended.

QualIntel OS structures your qualitative analysis from research question to submission package — surfacing candidate evidence for your review, never making decisions for you. Every coding choice, theme, and synthesis narrative stays yours.

No credit card required. Free for your first project.

Voyage AI semantic searchClaude Sonnet AI assistanceAPA 7 · Harvard · Chicago · VancouverOwen's (1984) criteria built-inFull audit trail export

The problem

Qualitative research is rigorous. The tools you're using aren't.

You're doing serious work — interviews, focus groups, document analysis — but managing it across Google Docs, sticky notes, and spreadsheets. When an assessor asks how you arrived at a theme, you scramble. The methodology is sound. The paper trail isn't.

Audit trails are an afterthought

You know every decision you made. But can you prove it to an examiner? Probably not — it's scattered across five apps and your memory.

Generic AI tools don't understand your method

ChatGPT doesn't know your codebook, your research questions, or what saturation means. It generates plausible-sounding output you can't defend.

Synthesis without structure is a risk

Writing theme narratives without checking quote diversity, RQ alignment, or single-voice dominance means surprises at examination. Late.

How it works

Six steps. One complete submission package.

A structured workflow that takes you from raw data to an assessor-ready package — without letting AI make analytical decisions on your behalf.

01Research design upload

Ground the system in your research

Upload your proposal, interview guide, literature framework, or rubric. QualIntel reads your research design before touching your data — so AI assistance is anchored to your methodology, not generic patterns.

02Codebook editor

Build your codebook

Create codes and themes grounded in your methodology — deductive from your literature, inductive from your data, or both. Every code is tied to your research questions. The AI can suggest initial codes from your design documents, which you accept or reject.

03Transcript processing

Upload your transcripts

Interviews, focus groups, field notes, or observations. Every segment is embedded with Voyage AI and indexed against your codebook — ready for semantic retrieval.

04Evidence review

Review candidate evidence

QualIntel surfaces candidate evidence segments semantically matched to each code. You accept or reject every suggestion. Nothing is coded without your explicit confirmation — always human-in-the-loop, never autopilot.

05Synthesis + Quality Checker

Write and quality-check your synthesis

Write narrative synthesis for each theme in a structured editor. Then run the built-in Quality Checker — which flags single-voice over-reliance, absent codes, RQ alignment gaps, and analytical modesty before you export.

06Submission package

Download your submission package

Export a complete ZIP: Evidence Pack DOCX, Codebook export, AI Disclosure Statement, Reflexivity Template, and a How-To Guide for your methods appendix. Formatted in APA 7, Harvard, Chicago, or Vancouver.

Quality Checker

A methodological review before you submit — not after.

The built-in Quality Checker reviews each theme narrative against objective data and interpretive criteria. It flags issues while you can still fix them.

Objective checks

  • ·Participant diversity — flags single-voice dominance above 60%
  • ·Absent codes — surfaces codes with no accepted evidence
  • ·Quote overreliance — warns if direct quotes exceed 45% of word count

Interpretive checks

  • ·RQ alignment — does the narrative address your stated research question?
  • ·Evidence grounding — are claims tied to specific participant data?
  • ·Analytical modesty — researcher voice versus AI-sounding language
  • ·A priori / emergent distinction — are code origins differentiated?

RQ Alignment

High confidence

Narrative directly addresses Primary RQ

Participant diversity

Objective

P3 accounts for 62% of cited evidence — consider broadening

Analytical modesty

High confidence

Researcher voice maintained throughout

A priori / emergent distinction

Medium confidence

Emergent codes not differentiated from codebook codes

Evidence grounding

High confidence

Claims tied to specific participant passages

QualIntel OS flags methodological considerations. Final analytical judgement stays with the researcher.

Submission_Package.zip

📄

Evidence_Pack.docx

Themes, audit trail, excerpts — formatted in your citation style

📋

Codebook.docx

All codes, descriptions, and RQ links

📝

AI_Disclosure_Statement.txt

Researcher-led framing of how AI assisted your analysis

🔍

Reflexivity_Template.txt

Structured prompts for your positionality and decision log

📚

How_To_Use_This_Package.txt

Guide for methods appendix and examiner review

Optional: Owen's (1984) recurrence, repetition, and forcefulness criteria applied to disclosure

Submission Package

Everything an examiner needs to evaluate your method.

One click generates a complete submission ZIP — not just a report. Your evidence pack, codebook, AI disclosure, reflexivity prompts, and a how-to guide for the examiner, all in one place.

  • AI disclosure framed around researcher-led interpretation — not generic boilerplate
  • Owen's (1984) criteria optionally applied to theme robustness assessment
  • Reflexivity template structured around your specific coding decisions
  • Citation-formatted in APA 7, APA 6, Harvard, Chicago, or Vancouver

Research Assistant

Ask questions across all your data. Get answers in evidence.

Stop rereading transcripts to find a single quote. The built-in Research Assistant retrieves across every document in your project — every answer cites the exact document and passage it came from.

  • Ask in plain languageJust ask which participants said what. The system retrieves — you interpret.
  • Evidence-linked answersEvery response cites document, page, and participant. Click to jump directly there.
  • Stays within your codebookThe assistant knows your codes, themes, and RQs — answers within your analytical frame.
  • Methodology-awareUnderstands thematic analysis, IPA, grounded theory, and framework method.
You

Which participants described barriers to accessing mental health support?

AI

Four participants described access barriers. Most common: cost and waitlists.

Interview_03.pdf · p.7 · P3

"I looked into it but the waitlist was six months. I just gave up after that."

Interview_07.pdf · p.4 · P7

"The cost was completely out of reach on a student income."

+ 2 more passages · Coded: barrier_access, barrier_cost

Built for rigour

Trustworthiness isn't a setting. It's the architecture.

Credibility, transferability, dependability, confirmability — the four pillars of qualitative trustworthiness, designed into every workflow.

Complete audit trail

Every coding decision, every theme revision, every AI suggestion you accepted or rejected — logged with timestamp, rationale, and researcher attribution. Your examiner traces every finding back to raw data.

Researcher-led, always

QualIntel surfaces candidate evidence segments. You accept or reject every one. Thematic interpretation, narrative synthesis, and analytical conclusions are yours — the system organises, retrieves, and audits.

Methodology-aware scaffolding

The platform understands thematic analysis, grounded theory, content analysis, and IPA — not just 'text analysis.' The synthesis editor, quality checker, and codebook tools all speak your discipline's language.

Assessor-ready documentation

Formatted evidence packs, codebooks, AI disclosures, and reflexivity templates — structured methodological documentation that demonstrates rigour at every level, not just at the finding.

Who it's for

Built for people who take qualitative methodology seriously.

Postgraduate researchers

Masters and PhD candidates running thesis research who need to demonstrate methodological rigour to supervisors and examiners — not just produce findings. The audit trail is your evidence of process.

Thesis research · Dissertation projects

Research consultants

Independent consultants and boutique agencies delivering qualitative insights to clients who want evidence-based conclusions. The submission package gives clients documentation they can actually interrogate.

Client reports · Stakeholder research

Evaluators and assessors

Programme evaluators, nonprofit impact teams, and education researchers who need defensible evidence for funders, boards, and regulatory bodies. The quality checker surfaces issues before the report goes out.

Programme evaluation · Impact assessment

Why QualIntel

The gap ChatGPT and NVivo both leave.

Based on standard workflows as of May 2026.

CapabilityGeneric AINVivo / Atlas.tiQualIntel OS
Methodology-anchored AI coding
Automated audit trail
Pre-submission quality checker
Submission package with AI disclosure
Structured codebook + RQ links
Human-in-the-loop confirmation
Evidence-linked Research Assistant
Cloud-native, any device
Priced for individual researchers

Pricing

Priced for researchers, not enterprises.

Start free. Pay when you need more.

Project
NZD 149
per project

One complete project from data upload to submission package export.

  • Unlimited documents
  • Full codebook + audit trail
  • AI-assisted coding
  • Quality Checker
  • Submission Package ZIP
  • 30 days access
Start a project
Most popular
Researcher
NZD 299
per month

For active researchers running multiple projects simultaneously.

  • Unlimited projects
  • Team collaboration (up to 5)
  • Priority AI processing
  • Version history
  • Email support
Start free trial
Institution
NZD 8,000
per semester

For universities, research units, and consulting firms needing platform access for cohorts.

  • Unlimited researchers
  • LMS integration
  • Dedicated onboarding
  • Custom methodology templates
  • Quarterly methodology review
Talk to us

Your interpretation. Methodologically defended.

Qualitative research is under scrutiny. Funders want evidence. Journals want transparency. Examiners want to trace every finding back to raw data. QualIntel OS makes that possible — without adding hours to your workflow or letting AI make decisions for you.

Start your first project free

No credit card required.