Define your codes once. FieldNotes applies them across every transcript — deductively, consistently, at scale. You review every decision. Full scientific rigour, without the manual grind.
Browser-based · No installation · GDPR-compliant · EU-hosted
Interviewer: How did you decide when to go back to work?
Participant: Honestly, it wasn't really my decision. My manager kept emailing during leave like nothing had changed, so I felt like I had to come back early just to keep my role.
Matches an existing code in your codebook — used in 4 other excerpts.
Build your codebook your way — then let the AI do the repetitive application work while you stay in control of every decision.
Upload interview transcripts in .txt, .md, .docx, or .pdf format. FieldNotes segments them into speaker turns automatically.
A large language model reads each excerpt and proposes a code — drawn from your codebook. The more codes you define, the more precisely it applies them. It can also flag when something new doesn't fit.
Review each proposal one by one. Approve it, merge it into an existing code, or reject it. Every decision is recorded — giving you a full audit trail for your methods section.
Not a generic AI tool dressed up as research software. Every feature reflects the actual qualitative coding workflow.
The AI applies your codes to new material, not the other way around. Define your framework once; FieldNotes codes consistently from it across every transcript. You approve, reject, or merge — never blindly accept.
Core featureYour codebook evolves as you code. Rename codes, merge duplicates, track definitions. FieldNotes warns you when the AI proposes something similar to an existing code.
CodebookKnow when to stop. FieldNotes tracks how often proposals generate new codes versus reusing existing ones — giving you an empirical signal for theoretical saturation.
AnalysisWrite reflections and methodological observations alongside your coding. Field notes are timestamped, searchable, and exportable — an essential part of rigorous qualitative work.
ReflexivityExport your full dataset as a structured spreadsheet: codebook definitions, coded excerpts with speaker attribution, and a complete decision audit trail.
ExportCollaborate with co-coders in shared projects. Role-based access keeps your data structured. Organisations can manage multiple projects and invite members across studies.
CollaborationEvery AI run archives the exact prompt text sent to the model. The audit page shows how each run's suggestions held up under your review — approved, merged, rejected — per run and per model. The disclosure detail AI-reporting standards now ask for.
New · TransparencyPeer-reviewed evaluations of generative AI in qualitative analysis show where it genuinely helps — and where it fails. FieldNotes is built to use the first and design out the second.
| What published AI-reporting standards ask for | How FieldNotes provides it |
|---|---|
| Which model and version was used | Logged on every run and every suggestion, included in exports |
| Full archived prompt text | The AI audit page stores the exact prompts sent to the model, per run |
| AI outputs logged before human revision | Proposals are append-only — never edited, never deleted |
| Verification of AI output quality | Acceptance rates per run and per model, computed from your own review decisions |
| Human final interpretive authority | The AI can only propose; every decision records the researcher who made it |
| Documented AI errors and remediation | Rejected proposals and review notes stay in the audit trail permanently |
The same evidence is equally clear about where today's models fail: culturally and emotionally nuanced interpretation, and quoting participants faithfully — published evaluations found models altering wording and splicing statements from different participants. FieldNotes is designed so those failure modes cannot corrupt your data. The model never generates quote text: excerpts are segmented verbatim from your transcripts before the AI ever sees them, and it can only attach suggestions to them. And nothing becomes part of your analysis without your explicit decision.
Reading: Gupta P, Topaz M, Connell KA, Yu H, Peltonen LM. Generative AI in qualitative health research: what the emerging evidence shows. Nursing Inquiry. 2026;33(3):e70140. · Lazarus MD, Zhao L, Gibson A, Martinez-Maldonado R, Stephens GC. Risky or rigorous? Developing trustworthiness criteria for AI-supported qualitative data analysis. Anatomical Sciences Education. 2026;19(2):330–337.
PhD students, postdocs, and PIs running interview-based studies. FieldNotes applies your theoretical framework consistently across transcripts and produces a citable audit trail.
HR teams, market researchers, policy analysts, and consultancies who need to systematically code interview data, focus group transcripts, or open-ended survey responses.
The same human-in-the-loop coding workflow, extended to visual data. If you work with images, recordings, or multimedia research data, email us to be first to know.
Nobody else gives you this. Copy it directly into your paper.
"Qualitative coding was conducted using FieldNotes (Unpinned B.V., 2025), an AI-assisted qualitative analysis platform. The software segmented interview transcripts into speaker turns and generated code proposals using a large language model. Each proposal was reviewed independently by the researcher, who approved, merged into an existing code, or rejected the suggestion. No codes were accepted without explicit human review. The full prompt text submitted to the model was archived for each analysis run, and per-run acceptance rates of AI proposals were recorded. The complete decision log, including all approved, merged, and rejected proposals, is available as a supplementary file."
NVivo and ATLAS.ti are powerful — and expensive, complex, and desktop-only. FieldNotes is built for how researchers actually work today.
| Feature | FieldNotes | NVivo | ATLAS.ti |
|---|---|---|---|
| Pricing (individual) | From €9.99/month | ~€500–600/year | ~€400–500/year |
| Browser-based (no install) | ✓ | ✗ | ✗ |
| AI code proposals | ✓ | Add-on | Add-on |
| Codes from your codebook (deductive) | ✓ | ✗ | ✗ |
| Human-in-the-loop review | ✓ | ✗ | ✗ |
| Saturation tracking | ✓ | ✗ | ✗ |
| Full audit trail | ✓ | Partial | Partial |
| Archived AI prompts & acceptance rates | ✓ | ✗ | ✗ |
| GDPR-compliant & EU-hosted | ✓ | Varies | Varies |
| Citable methods snippet | ✓ | ✗ | ✗ |
| Team collaboration | ✓ | ✓ | ✓ |
Considering ATLAS.ti? Read what changed after the Lumivero acquisition →
All plans include a 7-day free trial. A card is required to start, but you won't be charged until the trial ends.
Student and Researcher prices include VAT. Lab pricing excludes VAT (21% added at checkout). Student pricing requires a valid institutional email address.
Start your 7-day free trial today — card required, no charge until it ends. Cancel any time.
Start free trialQuestions first? Email us at azadali@unpinned.nl