CLARA
New
Any channel, any language. One complete picture.
CLARA is conversational intake. Someone reports in their own words, on whatever channel suits them, and CLARA turns it into one structured, decision-ready record.
An inbound phone call arrives in Japanese. CLARA detects the language and translates it to English live, recognises the caller and pulls through their record from your files, and fuses what is said with what is on file into one structured incident record with severity, enrichment, a confidence score and a recommended action.
A forklift dropped a load. No one is hurt.
Thank you. Where on site did it happen?
Near Bay 3. I will send a photo.
Illustrative. Example values are placeholders. Japanese is shown as one example; CLARA detects and translates the reporter’s language and keeps the original on the record.
The agents behind CLARA
Each one adds a layer the reporter never has to type
CLARA is one orchestrating agent, not nine separate ones. These are the nine responsibilities it carries as it listens, identifies, enriches and routes. All are illustrative and configurable.
Reporter identity
Who is reporting, and how we know.
Cover and account
Which account, policy or scheme the notification belongs to.
Triage
What happened and how serious, in your language and your categories.
Location
Where and when it happened, resolved from the conversation to a findable place.
Evidence capture
Photos and documents from any channel, with no app to install.
Evidence reading
What the supplied evidence actually shows, described in one factual line.
Witnesses
Who else may be able to supply evidence, and how to reach them.
Evidence position
What is established, what is not, and what is still missing. It stops before judgement.
Escalation
The right person told when attention is needed.
Why it is different
Three things a form cannot do
Any channel
Meet people where they are
Phone, email, text, video or in-app. The conversation arrives however suits the reporter, and ends up in the same place.
Any language
Speak freely, report clearly
CLARA detects the language and translates live, then standardises the record to your language while keeping the original for the audit trail.
Context aware
The record starts half-built
It recognises the reporter and pulls through what is already on file, then fuses that with what is said now. Two sources, one picture.
Capture, then intelligence
CLARA captures it. Logincident AI makes sense of it.
CLARA turns a messy, multimodal report into one clean record. Logincident AI then reads across every record, surfaces the patterns, and drafts the next step. Two layers of the same idea: signal out of noise.
Explore Logincident AIFrequently asked questions
What is CLARA?
Conversational intake. Someone reports in their own words, by voice, text, email or web form, and CLARA turns that into one structured record. It detects the language and translates live, keeping the original words alongside. It removes the form as the barrier to reporting, which is the oldest problem in this field.
How is that different from a chatbot?
A chatbot has a conversation. CLARA produces a record. The output is a structured, decision-ready report with the fields a system and a person both need, not a transcript somebody has to read and re-type.
Is CLARA the same as Logincident AI?
No, and this is worth being clear about because the names invite the confusion. They are separate products, licensed separately. CLARA is intake. Logincident AI is analysis of records that already exist. Neither depends on the other.
Can we see CLARA working?
Yes. CLARA is shown live in demonstrations, taking a report by voice or text and building the record as the conversation goes. It is licensed separately, and the team that has shipped the Logincident platform since 2012 is ready to take it to production with the first partner who wants it.
Related reading
Notes from the working world.
Is AI marking its own homework?
This week PHP2Shell caused mayhem around the world, and AI helped find it. So is AI marking its own homework, and who marks the marker?
23 Oct 2026 · 3 min read
Read article →
Somebody used to do this for a living
Jordan Nugent on the work that automation leaves behind: the compensating, connecting and cleaning that sits in no plan and on nobody’s job description.
19 Sep 2026 · 11 min read
Read article →See CLARA take a report
Book a walkthrough and watch a messy, real-world report become one structured record.
Book a demo