Platform Overview

The literature tells you
what exists.
Lacuna shows what doesn't.

Upload your corpus. In seconds, discover the concepts least covered, the bridges between ideas, and the white space where original contribution lives — entirely in your browser. No upload. No server. No limits.

Zero capture Deterministic Reproducible Auditable No file size limit
Start free → View plans
Privacy · Architecture

Your documents never leave
your device. By architecture.

This is not a privacy promise in a terms of service. It is an architectural fact. Lacuna processes everything locally in your browser. Our servers never see your files.

● Zero capture — verified by design
Every PDF, every TXT, every BibTeX file is processed entirely in memory on your own machine. The analysis engine runs client-side. Your API keys are stored only in your browser and used directly — they never pass through our servers. When you close the tab, the data is gone. We have nothing to retain, sell, or hand over.
Lacuna Labs
✔Documents processed entirely in your browser
✔No file ever sent to any server
✔API keys stored locally — never transmitted
✔Compatible with confidential and classified material
✔GDPR UK compliant by architecture, not policy
✔Same corpus, same parameters — same result, always
AI-based document tools
✗Every file uploaded to external servers
✗Content may be used for model training
✗Hard limits: typically 5–20 files per session
✗Not suitable for legally privileged documents
✗Subject to provider retention and jurisdiction policies
✗Results vary between runs — probabilistic
Capability AI assistants Document AI tools Lacuna Labs
Documents per session5 – 20 files10 – 50 files20 to unlimited *
File size limit10 – 25 MB eachVaries, often 10 MBBrowser memory only
Cross-document gap detectionNot availableNot availableCore feature — all plans
Documents sent to serversAlwaysAlwaysNever — any plan
Deterministic resultsNo — probabilisticPartialYes — same input, same output
* Trial: 20 documents · Scope / Education: 100 · Signal: 500 · Core: unlimited. File size bounded only by device memory — not our servers. AI tool figures based on publicly documented limits as of 2026. Lacuna Labs does not name or endorse any competitor.
01 · Corpus Analysis

Heat Map

Every document × every term. Central presence is shown at full intensity — peripheral mentions at reduced opacity. Coverage percentage per column reveals which concepts are systematically treated and which are gaps.

Lacuna Labs Heat Map — corpus analysis view
02 · Configuration

Settings & Domain filters

Calibrate the analysis to your field. Domain stopwords remove vocabulary so ubiquitous in a discipline that it carries no analytical signal — leaving only terms that genuinely differentiate documents.

Domain stopwords selector
Domain selector
Field-specific stopword filters — 6 superareas, dozens of subdomains
Analysis settings panel
Settings panel
Max terms, coverage threshold, word length, toggles for stopwords and case
03 · Corpus Analysis

Timeline

How each term peaked, declined, or emerged across the years inside your literature. Different from Pulse — Timeline measures what happened inside your documents over time, not what is happening in the world right now.

Timeline — term frequency over years
Timeline — document date configuration
Step 1 — document dates
Dates detected automatically from filenames, metadata and document text — overridable per document
04 · Corpus Analysis

Clusters

Two views of the same thematic structure. The adjacency network shows which documents share vocabulary — isolated nodes are outliers, often the most novel contribution angles. The membership table shows how strongly each document belongs to each identified theme.

Clusters — adjacency network view
Network view
Documents as nodes — lines = shared vocabulary above threshold. Isolated nodes = outliers.
Clusters — membership table
Membership table
Colour intensity shows how strongly each document belongs to each theme cluster
Clusters — theme composition
Theme composition
The vocabulary that defines each cluster — generated by AI Advisor or automatic grouping
05 · Intelligence

Pulse

Is the world moving toward your topic? Three independent signals — news media, academic publishing, patent filings — measured simultaneously. A topic silent in the press but accelerating in patents is pre-competitive intelligence.

Pulse Editorial — news coverage velocity
Trend Discovery — global press monitoring
Trend Discovery
57 RSS feeds · 12 regions · terms gaining momentum right now · updated 4× per day
Pulse Research — academic publication growth
Pulse Research
OpenAlex · 250M+ academic works · publication volume per year with trend signal
All features

One corpus. Ten analytical lenses.

Every tool runs on the same deterministic engine. No AI hallucinations. No black boxes. Same corpus, same parameters, same result — always.

01 · Corpus
Heat Map
Every document × every term. Central vs peripheral presence. Coverage per column.
02 · Corpus
Gap Analysis
Terms ranked by underrepresentation. Low coverage = opportunity for original contribution.
03 · Corpus
Bridge Map
Concepts shared across multiple documents — connective tissue between research streams.
04 · Corpus
Freq × Coverage
Two-dimensional map: how often vs how broadly. Core, background, specialist, emerging.
05 · Corpus
Timeline
How each term peaked, declined, or emerged across years inside your literature.
06 · Corpus
Clusters
Documents by thematic similarity. Outliers often signal the most novel angles.
07 · Intelligence
Pulse
News media, academic publications, patent filings — three independent signals simultaneously.
08 · Research
Author Intelligence
h-index, citations, ORCID, top papers, country — who is doing the work you haven't done?
09 · Research
Patent Explorer
Filing acceleration across Lens.org, EPO, PatentsView. Pre-competitive signal.
Use cases

Built for every kind
of serious inquiry.

From a PhD student mapping a dissertation to an intelligence team tracking a field — the same engine, calibrated to each context.

Academic · Scope / Education
PhD candidate mapping a dissertation proposal
Uploads 40 papers from the literature review. In 90 seconds, identifies 3 concepts with coverage below 20% — areas the field has discussed but never resolved.
→ Committee approved the gap as "genuinely novel contribution"
Corporate R&D · Signal
Innovation team benchmarking a technology landscape
Uploads 120 industry reports. Pulse shows a competitor field accelerating 280% in 2 years — invisible in internal briefings. No document left the company network.
→ Board presentation revised with pre-competitive signal
Investigative · Core
Journalist analysing 5 years of regulatory filings
Identifies systematic patterns across 800 documents. Results are fully deterministic — the same analysis run by any other researcher produces identical output.
→ Investigation cited the methodology as independently verifiable
Pricing

Start free.
Scale when ready.

Document and term limits depend on your plan. Zero capture applies on every plan — no document ever reaches our servers, regardless of tier.

Trial
£0
No time limit
20 documents · 50 terms
Heat Map + Gap Analysis
Bridge Map + Word Cloud
Compare 2 corpora
Pulse Intelligence
Timeline + Clusters
Author + Patents
CSV export only
Scope
£14.99
per user / month
100 documents · 200 terms
Full corpus analysis suite
Timeline + Clusters
Pulse Editorial + Research
Domain stopwords
Compare 3 corpora
CSV + HTML + PNG export
Author Intelligence
Patent Explorer
Education
£499
per institution / month
All users on verified domain
Same capabilities as Scope
100 documents · 200 terms per user
Auto-access by institutional email
Universities + research institutes
Pulse Editorial + Research
CSV + HTML + PNG export
Author Intelligence
Patent Explorer
★ Most popular
Signal
£249
per organisation / month · 25 users
500 documents · 500 terms
Full corpus analysis suite
Full Pulse (Editorial + Research + Patents)
Author Intelligence
Patent Explorer
Compare 5 corpora
PDF export
Team + group management
Core
from £1,499
per month — contact us
Unlimited documents + terms
Everything in Signal
Investigative language module
CDR pattern analysis
Compare 10 corpora
JSON export + full data access
Self-hosted secrets store
SSO + audit log
Dedicated support
Zero capture on every plan — including Trial
Documents are processed entirely in your browser on every plan. No file reaches our servers. Compatible with classified information, legally privileged material, and trade secrets.

Ready to see what your
corpus is missing?

No installation. No upload. No credit card to start. Open the platform and drop your first documents in 30 seconds.

Start free at lacunalabs.eu →