The familiar interface
Chats · B. Reyes's phone
4 CONVERSATIONS · 2 WITH HITS
cash
AT
3 HITS
Alice Tan · Vendor Rep
Split it across the two cash pay…
PS
1 HIT
Procurement Squad
Voice note · 0:41
CK
—
Carlos K.
Photo · site_visit.jpg
MF
—
Mum ❤
Call me when you land
AT
3 HITS
Alice Tan · Vendor Rep
FROM PHONE EXTRACTION · CUSTODIAN B. REYES
TUE, 14 MAR 2023
Invoice cleared. Same cash arrangement as last time?
Keep it off email. I'll call you.
DELETED BY SENDER
This message was deleted.
RECOVERED FROM DELETED DATA
Split it across the two cash payments, nothing in writing.
receipt_scan.jpg
IMAGE · OCR'D · TEXT SEARCHABLE
One continuous conversation — not a pile of per-day chat documents.
Pick a chat, read it whole
Participants, replies, media and voice notes, in order, across every day it ran.
Search lands you in context
You arrive on the message with the exchange around it intact. Images inside chats are OCR’d.
Day slices, one click away
The day-slice stays the coded unit — defensible and countable — and reviewers can switch to it at any time.
Built for forensic extractions
Cellebrite UFDR, Android
Decoded chats straight from the report — identifiers, device timestamps, participants, recoverable deleted rows.
iPhone extractions
The WhatsApp database as the device stored it: every message type, media, voice notes, reactions, deletions.
Never a screenshot pile
Provenance travels with every message, so the exhibit is the extraction — not someone’s retyping of it.
Can you recover deleted WhatsApp messages?
AT
Alice Tan · Vendor Rep
FROM PHONE EXTRACTION · CUSTODIAN B. REYES
TUE, 14 MAR 2023
Invoice cleared. Same arrangement as last time?
Keep it off email. I’ll call you.
⊖ DELETED BY SENDER
This message was deleted.
2
↻ RECOVERED FROM DELETED DATA
Split it across the two cash payments, nothing in writing.
1
1Certified deleted. The forensic tool found the row in the device database. The text usually survives.
2Claimed deleted. The app’s own tombstone: the sender recalled it. The marker never says what it said.
3Absent. The unmarked messages: nothing in the extraction says they were deleted, and the record says so.
Every state travels with the message — through search, tagging, batching and export — so what you put in front of a tribunal says exactly what the source supports.
What survives a courtroom
Provenance
Which device or account, produced by whom, when, with what.
Integrity
A hash at ingestion: the exhibit is provably the item collected.
Completeness
What the artefact covers, on the record.
An auditable review
Who looked at what, and what they decided.
Reproducibility
The same source yields the same output, so an opponent can follow it.
How Kendraa handles it
Ingestion
Cellebrite UFDR and iPhone extractions, with media, voice notes and deleted rows.
Reconstruction
Participants, timestamps, replies, attachments and voice notes, rebuilt into conversations.
Search
Plain English across chats and documents. No query syntax, no operator.
Deployment
Your own AWS account, in India. The AI runs in the same environment.
See the security posture, or write to us about an artefact you are stuck on.