Loading the anonymous daily summary…
Invitation App / Public research record
Database
Updates
A live, interactive view of database totals and reviewed reports—showing what the app recorded and what still needs analysis.
Resurrections · Received records and published reviews
Paranormal Invites™
Resurrections
Open the latest received session to explore its recorded words, question outcomes, and supporting measurements. Published editorial reviews remain separately dated; internal thresholds and selection mechanics stay private.
Open the latest received session →Live intake snapshot ·
The received record
updates automatically.
This snapshot is built from anonymous, date-level totals exposed by the public database summary. It refreshes automatically while the page is open. Detailed case reviews are separate and can have a later review date.
Totals describe what the app recorded; they do not establish a cause.
Automatic public summary
Live intake totals
These anonymous totals refresh from the central database on page open and every 15 seconds. Tap a total to filter the dated entries below; tap a date to open its counts and individual sessions. Open a session to inspect its words and supporting readings.
Waiting for the database…
Open a date, then choose a session to see recorded outputs and supporting measurements. Participant identities, exact locations, private notes, question audio, raw downloads, and selection rules stay private.
Open a real submitted session review — Explore EMF words, questions, phonetic fragments, SLS, and EVP candidates in plain language.
All accepted uploads combined
The current database trail
Daily rows retain the database reporting dates. Individual submission times use Pacific Time. Each row combines that reporting day’s submissions. Expand a date, then open an individual session. Counts describe app records, not confirmed findings.
Showing every date.
Word outputs
Recorded word items in the public date-level summary.
EVP candidates
Candidate counts recorded in the public date-level summary.
SLS items
Item counts recorded in the public date-level summary.
Reading these totals: Intake counts refresh automatically from the public daily summary. Individual sessions show a limited public record of words, question outcomes, and measured readings. Participant identities, raw files, private notes, and selection rules stay private; these records do not establish a cause.
All contributors combined
A day-by-day
database trail.
Each date combines anonymous submissions and their linked analysis. A trigger records an app condition; it does not establish what caused it.
October 4, 2026
8 submissions106 word outputs · 0 EVP candidates · 0 SLS items · 1 detailed analysis
One analyzed session recorded 6 question prompts and 4 responses. The four question-mode words were TEXTILE, RUST, NASAL, and CONDUCT; COMMERCE was recorded in Live Word mode. The report logged EMF sensor events before outputs; movement was not listed as an output channel. This device had no ambient-temperature sensor.
Phonetic mode: “chul ba” · “pat net doth yail” · “fur feesh leesh” · “sak yaek sos” · “nee tath vaet” · “baes thik.” These are app-generated syllable strings, not captured speech or a translated sentence. Seven same-day submissions do not yet have linked detailed analyses.
October 3, 2026
5 submissions5 word outputs · 0 EVP candidates · 0 SLS items · 0 detailed analyses
The intake counters report five word outputs across the day’s submissions. No detailed analysis is linked yet, so the exact words, question responses, and trigger context are not shown.
September 30, 2026
2 submissions2 word outputs · 0 EVP candidates · 11 SLS items · 2 detailed analyses
The analyzed records list ABOVE and DISTRICT as Live Word outputs; 4 separate question prompts had 0 recorded responses. The SLS figure is an intake item count, not a count of publicly approved videos. EMF was available in the analyzed device records; ambient temperature was not.
September 29, 2026
2 submissions0 word outputs · 0 EVP candidates · 1 SLS item · 2 detailed analyses
Across the analyzed records, 1 question prompt produced no recorded response. The SLS item remains a private submission count; its underlying media has not been published here.
September 28, 2026
1 submission0 word outputs · 0 EVP candidates · 0 SLS items · 1 detailed analysis
The analyzed session logged 2 question prompts and no recorded answers. The quiet result remains part of the record.
September 25, 2026
7 submissions1 word output · 1 EVP candidate · 0 SLS items · 7 detailed analyses
Across these records, 7 question prompts had 1 recorded response: BOND. Its analysis logged an EMF-associated app response; the report does not establish what caused the sensor event. One EVP candidate was counted at intake and remains a candidate, not verified speech.
September 19, 2026
10 submissions58 word outputs · 1 EVP candidate · 24 SLS items · 0 detailed analyses
These are combined intake counters for the day. No detailed analyses are linked yet, so individual words, responses, or event context are not available for this public summary.
September 18, 2026
3 submissions11 word outputs · 0 EVP candidates · 1 SLS item · 0 detailed analyses
Combined intake counters only; the underlying session-level words and trigger context still need detailed analysis.
September 17, 2026
1 submission0 word outputs · 0 EVP candidates · 0 SLS items · 0 detailed analyses
A submitted quiet session is included in the database total.
September 16, 2026
4 submissions15 word outputs · 1 EVP candidate · 5 SLS items · 0 detailed analyses
Combined intake counters only; detailed word and trigger context is not yet linked.
September 15, 2026
1 submission0 word outputs · 0 EVP candidates · 0 SLS items · 0 detailed analyses
A submitted session with no word, EVP-candidate, or SLS items in its intake counters.
September 14, 2026
6 submissions1 word output · 1 EVP candidate · 5 SLS items · 0 detailed analyses
Combined intake counters only; individual media and event context remain private pending detailed review.
September 13, 2026
12 submissions22 word outputs · 3 EVP candidates · 0 SLS items · 0 detailed analyses
These are the day’s combined intake totals. Exact word content and EVP review are not available in the linked detailed-analysis records yet.
September 12, 2026
7 submissions74 word outputs · 0 EVP candidates · 8 SLS items · 0 detailed analyses
Combined intake counters only; individual word sequences and sensor context still need detailed analysis.
September 9, 2026
3 submissions4 word outputs · 0 EVP candidates · 3 SLS items · 0 detailed analyses
Combined intake counters only; exact words and context are not inferred from counts.
September 7, 2026
1 submission4 word outputs · 0 EVP candidates · 0 SLS items · 0 detailed analyses
The intake summary records four word outputs; the words and trigger context are not yet available in a linked analysis.
September 6, 2026
5 submissions10 word outputs · 1 EVP candidate · 0 SLS items · 0 detailed analyses
Combined intake counters only; the EVP item remains a candidate and is not represented as confirmed speech.
September 3, 2026
2 submissions5 word outputs · 0 EVP candidates · 0 SLS items · 0 detailed analyses
The earliest research submissions currently in the database. Their aggregate counters are preserved while detailed analysis remains unavailable.
How to read this timeline: “Submissions” counts all contributors together for that date; it never identifies who sent a record. Intake counters are not independently verified findings. Detailed words and context are shown only for days with linked analysis. SLS item counts do not mean the corresponding videos are public; Mike reviews media separately before any publication.
Three levels of detail
A quick pulse.
A deeper review.
Choose the timeframe that fits what you want to know. A report is published only when its information has passed review.
The absence of a new report does not mean information is being hidden. It means the review is incomplete or there was no meaningful change to publish.
Weekly Field Note
A short, timely check-in when the database or testing program has something meaningful to report.
What it can include
- Permitted session packages received
- Items newly entering or completing review
- New contamination flags
- Noteworthy quiet or no-response sessions
- App build or testing changes
- A clear “no material change” note when appropriate
Weekly rule: early activity is reported as an observation or candidate—not a conclusion.
Biweekly Research Digest
A two-week comparison that gives possible patterns more context than a single field note can provide.
What it can include
- Submission and review totals for the period
- Repeated activity across separate sessions
- Research-mode comparisons
- Negative and no-response results
- Environmental or handling contamination
- Candidates promoted, downgraded, or left unresolved
Biweekly rule: a repeated pattern becomes more interesting, but repetition alone still does not identify its cause.
Monthly Database Review
The fullest public summary, combining the month’s verified counts, classifications, operating changes, and unanswered questions.
What it can include
- Total permitted submissions received
- Reviewed, pending, excluded, and unresolved records
- Results by research mode and classification
- Negative-result and contamination rates
- App versions and methodology changes
- Limits, follow-up tests, and next-month priorities
Monthly rule: every total must have a defined source, review state, and reporting window.
Real records, personal details removed
See what a database
intake summary looks like.
These historical examples come from Build 74 and Build 75 submissions received September 14–15, 2026. Intake totals and session records above load live. The historical examples below retain their original review dates. Examples show only limited facts needed for public understanding.
Exact session IDs, tester identity, file names, storage paths, security hashes, location details, and private archive contents are intentionally withheld.
Long quiet-session summary
- Build
- Prototype 8.0 · Build 75
- Duration
- 34 min 39.7 sec
- Words
- 0
- EVP / SLS
- 0 / 0
Zeroes in these three summary fields stay in the record. They do not prove every sensor was inactive; the private package must be reviewed before anything more is said.
Word-output summary
- Build
- Prototype 8.0 · Build 74
- Duration
- 12 min 59.0 sec
- Words
- 1
- EVP / SLS
- 0 / 0
The intake row confirms that one word output was recorded. It does not establish its relevance, meaning, or cause.
Multi-capture summary
- Build
- Prototype 8.0 · Build 74
- Duration
- 5 min 33.7 sec
- Words
- 0
- EVP / SLS
- 1 / 5
The intake row confirms one EVP capture and five SLS captures. Each item and its surrounding conditions still require private review before classification.
Numbers need context
More than
a running total.
A larger database does not automatically mean stronger evidence. Each public update shows how much information was reviewed, how it was classified, and what limits remain.
Submissions received
Only session packages deliberately sent with the green database button during the stated reporting window.
Review state
How many records are pending, reviewed, excluded, or still waiting for more context.
Classification
Which results are unresolved, possible correlation candidates, contamination likely, or preserved null sessions.
Negative results
No-output sessions are counted with their available sensor context. Below-threshold activity can then be compared with sessions that produced a word, phonetic sound, or other response.
App and method version
The operating build and research rules used during the window, including relevant changes.
Limits and next tests
What the information cannot establish and which controlled follow-up would make it more useful.
Review before release
The publication trail
The public page is a reviewed report—not a live window into private database records.
- 01
User submits
A participant deliberately sends the permitted session package.
- 02
Record is secured
The original research record is preserved for authorized review.
- 03
Identity is separated
Public reporting removes identifying information and avoids precise location disclosure.
- 04
Evidence is reviewed
Baselines, timing, movement, audio, radio state, and possible contamination are checked.
- 05
Results are classified
Active, quiet, contaminated, unresolved, and excluded records remain distinguishable.
- 06
Aggregate is published
Only verified totals, trends, limitations, and non-identifying examples reach the public update.
THE PRIVACY BARRIER REMAINS IN PLACE
Public progress.
Private participants.
Database updates count only session evidence a participant deliberately permits and submits. They do not publish precise GPS coordinates, contacts, messages, call history, passwords, browsing activity, unrelated phone files, or private raw media by default.
How the app produces and reviews outputs
A sensor change must qualify before an evidence word is produced. Waiting alone does not force a word, and network responses do not select words. Read the public explanation of timing, network use, movement checks, and what a session audit can establish.
How sensor-triggered outputs work