Open Beta — Free to use

Where interpretation
becomes evidence.

Build a shared analytical framework, apply it to any body of evidence, and produce results that are verifiable — not just claimed. Human-led. AI-assisted.

No credit card required · Free during Open Beta

app.lenselot.com/project/123/code
Row 3 of 47
6%

Source data

Title

Social media use and adolescent mental health outcomes

Author

Smith, J. et al.

Year

2024

Coding form

Sentiment *

Topic *

AI Suggestion

Negative

Grounded in

▸ Strauss & Corbin (1998)

▸ 6 similar coded rows

Codebook · v4

Positive12
Negative8
Neutral4

Used across research disciplines

Academic ResearchUX ResearchPolicy AnalysisMarket InsightsClinical ResearchInvestigative Journalism

From raw evidence to defensible findings.

No setup wizard. No implementation consultant.

01

Upload your evidence

Spreadsheets, PDFs, audio, video — bring the full body of material your analysis will rest on. Lenselot handles all three source types in a single project.

02

Build your interpretive framework

Define what you are looking for in terms your whole team can apply. Build a codebook from scratch, import an existing one, or let the AI suggest a starting structure from your data. You decide what stays.

03

Apply, verify, and publish

Code every unit. Measure inter-rater agreement automatically. Produce findings backed by a methodology that is versioned, transparent, and citable — not just described in a footnote.

Your framework, under version control.

Every change to your codebook is tracked, attributed, and reversible. Propose revisions, review diffs, merge with team sign-off. Your analytical framework has a history — and that history is part of your methodology.

  • Full version history with per-change attribution
  • Proposal and review workflow before merging changes
  • Diff view shows exactly what changed and why
  • Restore any previous version at any time
Codebook: Mental Health Studyv4 · current

Version history

v4

Added 'Resilience' code

Sarah K.

v3

Merged proposal #12 — 3 codes revised

James R.

v2

Refined 'Anxiety' definition

Priya M.

v1

Initial framework

Sarah K.

Change · Anxiety — definition

- “Expressed worry or fear”

+ “Reported symptoms including worry, fear, or avoidance behaviours”

Built for teams. Designed for disagreement.

Multiple coders. Configurable overlap. Automatic IRR. When disagreement surfaces — which it should — adjudication resolves it with a complete audit trail, not a silent override.

  • Concurrent multi-coder projects with row-level assignment
  • Cohen's κ and percent agreement computed per dimension
  • Calibration rounds to align the team before production coding
  • Adjudication queue shows all coder responses side by side

IRR Overview · Sentiment

κ = 0.84

Substantial

Percent agreement

91%

Coders

S

Sarah K.

142 rows

94%

J

James R.

142 rows

91%

P

Priya M.

142 rows

88%

3 rows need adjudication

Review →

Agentic AI that works from your data — not thin air.

Our grounded AI agents read your uploaded literature, cross-references previously coded rows, and proposes codes with explicit reasoning. Every suggestion cites its source. You accept, modify, or reject — always.

  • Suggestions grounded in your uploaded literature
  • Cross-references similar rows already coded by the team
  • Accept, modify, or reject — no silent auto-application
  • AI attribution tracked separately in the audit log

AI Assistant · Row 12 of 47

Evidence

“The intervention helped participants recognise triggers before escalation, reducing crisis episodes over 12 weeks.”

Suggested code

Psychoeducation

Teaching individuals to understand and manage their own mental health symptoms.

Grounded in

Strauss & Corbin (1998) — open coding method
Similar to 8 previously coded rows in this dataset

The analytical rigour your work deserves.

Built for teams who need their methodology to hold up — under peer review, under editorial scrutiny, under replication.

Dataset, document & multimedia

CSV, XLSX, PDF, audio, and video all work in the same project workflow.

Reusable codebooks

Build a codebook once and apply it across multiple projects. Four code types, each with a description shown at coding time.

Focused coding interface

Each row as a vertical card. Source fields stacked above the coding form. Navigate by keyboard or jump directly to any row.

Multi-coder workspaces

Invite collaborators, configure row overlap for reliability testing, and manage roles — all within one workspace.

Inter-rater reliability

Cohen's κ and percent agreement per dimension, computed automatically. Export IRR statistics for reporting or publication.

Export at any time

Download as .xlsx or .csv at any stage. All original columns, one column per code, plus row status. No lock-in.

Also included

AI codebook generation

Let AI suggest an initial framework from your data. You review every suggestion — nothing is applied automatically.

Evidence search

Semantic search over PDF corpora. Accepted passages are highlighted and tracked per row.

Calibration rounds

All coders label the same rows before the main run to align on the rubric.

Adjudication

Designated reviewer sees all coders' answers side by side and selects the authoritative code.

Wherever structured judgment meets unstructured evidence.

Any team that applies a framework to a body of material — repeatedly, verifiably, with others.

Academic Research

Systematic reviews, grounded theory, thematic analysis. Produce methodology sections that hold up to peer review.

UX Research

Interview transcripts, usability sessions, survey verbatims — structured into defensible findings teams can act on.

Policy Analysis

Government documents, public consultations, legislative records. Track how frameworks evolve across policy cycles.

Investigative Journalism

Collaborative coding of source materials, leaked documents, and structured datasets — with a full audit trail.

Clinical Research

Patient records, adverse event categorisation, trial documentation. Rigorous coding with IRR built in.

Market Insights

Focus groups, brand monitoring, competitor materials. Build a reusable framework that compounds across studies.

Simple pricing.

Free during Open Beta. Paid plans introduced after beta — early users notified in advance.

Open Beta

Free

during Open Beta

Everything included:

  • Unlimited projects
  • Dataset, document & multimedia
  • Reusable codebooks
  • Multi-coder workspaces
  • IRR & calibration
  • Export as Excel / CSV
  • AI code suggestions
Get started free

Paid plans will be introduced after the Open Beta. Early users will be notified in advance.

Frequently Asked Questions

Can't find an answer? Contact us for support.

Your research, amplified.

Free during Open Beta. No credit card required.

Get started free