Deterministic machine-learning evidence analysis
Turn thousands of text messages into evidence you can source and review.
Textimony runs a deterministic machine-learning pipeline over the export you import: it classifies behavioral signals, reconstructs chronology, measures patterns, and links every result back to the underlying messages.
Repeatable inputs. Auditable outputs. No chatbot guesswork.
Deterministic machine-learning pipeline
Deterministic classification. Chronological reconstruction. Evidence-linked results. Exact-data review. A sequence of trained models runs over the frozen case input: an issue classifier scores each message for candidate categories (harassment, threats, coercion, custody interference, boundary events), embedding models retrieve related messages by meaning, a cross-encoder reranks candidates by fit, and a span model marks the exact words behind each candidate. The dataset and analyzer configuration are hashed for each run so the same input and configuration reproduce the same output, and each result keeps its message, timestamp, participant, and case-run references.
Five-stage review workflow
Import a supported export, confirm participants and time handling, run analysis on the frozen case input, inspect candidate signals with source context, and download or rebuild the report after review.
Clearly labeled homepage example
The homepage chart and message excerpts are an illustrative sample, not real case data. They show how a review can progress without presenting the example as a customer result.
What analysis provides
Configured checks and classifier models surface candidate patterns. Candidates retain message, timestamp, participant, and case-run references for human review.
What remains a human decision
A reviewer confirms the imported record, reads the surrounding conversation, records inclusion or dismissal decisions, and decides what to share outside the workspace.
Supported record formats
Current import routes support iMessage or Android-compatible CSV, Android SMS XML, WhatsApp TXT, EML email, and JSON or JSONL records. The mapping step still requires the user to confirm participants and time handling.
Private case boundary
Private uploads, normalized messages, participant mapping, analysis runs, reviewer decisions, and report files live inside the signed-in case workflow.
Four subscription paths
Personal Organizer is $49 per month, Personal Analysis is $149 per month, Professional Case is $599 per month, and Firm Workspace is $999 per month. The first three are case scoped; Firm Workspace is account scoped.
Reports and manifests
Completed runs can provide PDF, CSV, JSON, and manifest downloads. Reviewer inclusion and dismissal decisions can be used to rebuild the report for that run.
Choose the next step
Start one case, read the methodology, review privacy and trust controls, or compare the account-wide plan for recurring case work.
Frequently asked questions
What does Textimony do?
Textimony organizes large text-message exports — SMS, iMessage, WhatsApp, and email — into a reviewable case. A deterministic machine-learning pipeline classifies behavioral signals, reconstructs chronology, measures patterns, and links every result back to the underlying messages so a person can review them in context.
Is Textimony a chatbot or an AI assistant?
No. Textimony is not a chatbot, copilot, or advice engine. It is a measurement and evidence-analysis workspace. The pipeline is deterministic, so the same messages and the same configuration reproduce the same classifications and the same chronology instead of a differently worded answer each time.
What does deterministic machine learning mean here?
Each run records a hash of the dataset and the analyzer configuration. Given the same input and configuration, the pipeline returns the same results, which makes the outputs repeatable and auditable rather than regenerated on every request.
Does Textimony decide guilt, intent, or legal conclusions?
No. Textimony surfaces candidate signals for a reviewer to inspect; it does not determine guilt, intent, truth, or any legal or psychological conclusion, and it does not claim perfect accuracy. A person confirms the record and decides what matters.
Which message formats can Textimony analyze?
Current import routes support iMessage or Android-compatible CSV, Android SMS XML, WhatsApp TXT, EML email, and JSON or JSONL records. During setup the user confirms participants and time handling before analysis runs.
Who is Textimony built for?
It is built for legal and professional evidence review — attorneys, investigators, and behavioral professionals — and for anyone who needs to turn a long message record into an organized, source-linked case they can review.