TRACE Ai · Aviation safety investigation software Hear every account. Keep the decision with investigators.

TRACE Ai is aviation safety investigation software that takes each investigation through the same eight-stage human-factors method, drafts each stage with its autopilot and gathers first-hand accounts. Tracy, its AI voice interviewer, phones each person with questions an investigator has approved, and the drafted analysis codes candidate factors against HFACS and Threat and Error Management, citing the interview lines behind each item. Investigators accept or reject each item and sign off the findings.

Email me the brochure, TRACE Ai
Safety investigators, safety managers and authorized reviewersIndependent productFits your operation as it changes
TRACE Ai: AI-Powered Aviation Safety InvestigationsPlay video: TRACE Ai: AI-Powered Aviation Safety Investigations
A 3-minute film of TRACE Ai: a safety report becomes a case, Tracy interviews the crew, and an investigator signs off. Synthetic demonstration data.
In brief

What is TRACE Ai?

TRACE Ai is FlightAtom's aviation safety investigation software. It guides each investigation through a structured eight-stage method, drafts the work of each stage, and uses Tracy, an AI voice interviewer, to phone the people involved with questions an investigator has approved. Every transcript is filed into the investigation's evidence, and findings and sign-off stay with the investigator.

Category
Aviation safety investigation software
Built for
Safety investigators, safety managers and authorized reviewers
Applied AI
Drafts each stage, proposes who to interview, runs consented AI phone interviews, and drafts cited analysis across the accounts; sign-off stays with people
Availability
Available · independent product
Fits your operation
As your operation grows or changes, we adapt TRACE Ai to fit, and you approve each change before it goes live.
Inside TRACE Ai

Walk through TRACE Ai, from safety report to sign-off.

Follow TRACE Ai from the safety report to sign-off: what the autopilot drafts, what an investigator approves, what Tracy asks, and what stays with people. Each step shows a current product screen, and the interview replays as text.

Current TRACE Ai screens with synthetic demonstration data and neutral presentation branding. Steps 1 to 4 and step 7 follow one fictional investigation, the go-around of flight DEM 214 at Demo Field; steps 5 and 6 show a second, a turbulence encounter on DEM 316. The people, interviews and analysis were written for the demonstration.

Step 1 of 7

The autopilot drafts each stage of the eight-stage method

Synthetic case · DEM 214 go-around

An investigator opens an investigation in TRACE Ai from a captain's safety report about an unstable approach and go-around, and starts the autopilot, which drafts each stage of the eight-stage method into the working notes, beginning with the plan and an evidence register, under a heading that asks for review before anyone relies on it. Then it waits for the people involved to be heard.

On this screen
  1. Fictional occurrence: DEM 214 at Demo Field
  2. Plan & prepare, drafted and tagged Autopilot draft
  3. Working notes with an evidence register, marked for review
  4. Waiting for you: nobody is contacted until you approve
TRACE Ai investigation page with synthetic demonstration data for a fictional unstable approach and go-around of flight DEM 214 at Demo Field. The investigation guide lists eight stages from Plan & prepare to Review & follow up: Plan & prepare is ticked and tagged Autopilot draft, and Collect & debrief is open with its purpose, working outputs and working notes headed Autopilot draft. Below, the Autopilot panel shows its eight-stage strip, the status Waiting for the interviews before analysing, and a Waiting for you note that one person's questions need approval before anyone is contacted.
Work the eight-stage methodDemonstration data · Click to enlarge
Step 2 of 7

Approve the questions before anyone is called

Synthetic case · DEM 214 go-around

At Collect & debrief, TRACE Ai's autopilot proposes the people whose first-hand accounts would close specific gaps, gives the reason for each, and drafts their questions from the investigation record, opening with an invitation to give their own account. Nobody is contacted until an investigator approves that person's exact question set. Approval locks the set to a fingerprint that is checked again before dialling, and the person picks a call time from an emailed link.

On this screen
  1. Why the first officer's account is needed
  2. Seven drafted questions, starting with their own account
  3. Approve, and the booking link is emailed
  4. Waiting for your approval
TRACE Ai Autopilot panel with synthetic demonstration data: the status Running, an eight-stage strip with Plan & prepare drafted and Collect & debrief in progress, a Waiting for you note, and People to interview. A synthetic captain is marked Interviewed with the transcript in the evidence. A synthetic first officer is marked Waiting for your approval, with seven drafted open questions and an Approve and email booking link button. A synthetic dispatcher is marked On the call. Each person has the reason their account is needed.
Approve before anyone is calledDemonstration data · Click to enlarge
Step 3 of 7

Consent first, then Tracy takes the captain's account

Synthetic case · DEM 214 go-around · captain's interview

When the call connects, the consent notice plays first, as fixed text the model cannot rephrase. After the captain consents, Tracy, TRACE Ai's AI voice interviewer, introduces herself as an AI and asks for the captain's account in their own words and in any order. She lets the captain finish, follows up on what they said, works through the approved questions and reads back a summary for the captain to correct. The transcript is filed into the investigation's evidence.

On this screen
  1. The consent notice, before anything else
  2. Tracy says she is an AI interviewer
  3. The captain's own account, uninterrupted
  4. Approved question set, locked by its fingerprint
TRACE Ai voice interview transcript with synthetic demonstration data. The consent notice says the interview is AI-assisted and transcribed and can be stopped by pressing 9, and the synthetic captain consents. Tracy (labelled SYSTEM) introduces herself as an AI interviewer and asks the captain to describe the approach and the go-around at the fictional Demo Field, in their own words and in any order. The captain answers over two turns without interruption. Tracy asks two follow-ups built on their words, then what they watched at a thousand feet. Beside it: the booking and the plan's provenance.
Consent first, then their accountDemonstration data · Click to enlarge
Transcript replay (text)The synthetic captain's interview on the screen above. TRACE Ai labels Tracy's lines System and keeps the transcript as text; the audio is not recorded.
  1. Tracy (AI)

    This is an AI-assisted safety interview that will be transcribed for authorized Safety investigators. You may stop or withdraw at any time, or press 9 at any time to stop and request human follow-up. Do you consent to continue?

  2. Captain (synthetic)

    Yes, I consent. Happy to go ahead.

  3. Tracy (AI)

    Thank you for agreeing. I'm Tracy, an AI interviewer from the Demo Airline safety department, following up on your report about the go-around at Demo Field on 7 October. This is a just-culture conversation, only about learning and prevention, never blame. In your own words and in any order you like, could you describe the approach and the go-around, until you were established in the climb?

  4. Captain (synthetic)

    Sure. We were vectored onto the ILS for runway two-seven with the first officer flying. Approach asked us to keep one-sixty knots to four miles, and by the time we were cleared to land we were already a bit fast and high. The tower was busy, and we were number two behind a turboprop that was slow on final.

  5. Captain (synthetic)

    The rain got heavier inside five miles. At a thousand feet we were still fifteen knots fast. I called speed, then go around at about six hundred feet, and he flew it straight away. We came round, briefed again for the same runway and landed off the second approach.

  6. Tracy (AI)

    You said you were already a bit fast and high. What was going through your mind at that point?

  7. Captain (synthetic)

    Honestly, I thought we'd get it back. The tailwind on the ATIS was only five knots and we'd briefed the wet-runway figures, so I was watching the speed come off rather than thinking about the gate.

  8. Tracy (AI)

    Mm, understood. You mentioned Approach asked you to keep one-sixty knots to four miles. What did you understand about that request, and how did it fit with the approach you had briefed?

  9. Captain (synthetic)

    We took it as normal sequencing. We'd briefed a flap-three approach with the wet-runway numbers, and one-sixty to four miles left us less room to slow down than we'd planned for. Neither of us said that out loud.

  10. Tracy (AI)

    Thanks, that detail helps. As you approached a thousand feet, what were you paying most attention to, and what did you have in front of you about the speed and the flight path?

Lines 11 to 18 continue the synthetic interview written for the demonstration, beyond the part on the screen, through the read-back and the captain's correction.

  1. Captain (synthetic)

    The speed, mostly. It was coming back, but we were still about fifteen fast and a little high on the glideslope. The automatic callout came, but neither of us said stable or unstable. That's really why I filed the report.

  2. Tracy (AI)

    Mm, I see. You said the rain got heavier inside five miles. How did the rain and the wet runway affect what you were seeing and expecting on short final?

  3. Captain (synthetic)

    The wipers were on high and the approach lights were smeared, but we had the runway. Knowing it was wet, I wanted us on speed at the threshold, so I kept waiting for the speed to come off instead of looking at the whole picture.

  4. Tracy (AI)

    Okay, I follow. You said he flew the go-around straight away. What happened in the flight deck when you called it, and how was it carried out?

  5. Captain (synthetic)

    I said go around, he pressed TOGA and called flaps, and I made the rest of the calls. It was clean. We told the tower and came round for a normal landing.

  6. Tracy (AI)

    Thank you for walking me through that. Let me read back what I've heard. You were asked to keep one-sixty knots to four miles, which left less room to slow down than you'd briefed, and you were fast and high at the landing clearance. The rain got heavier, you were mostly watching the speed, and at a thousand feet neither of you said stable or unstable. You called speed, then go around at about six hundred feet, the first officer flew it straight away, and you came round for a normal landing. Is that right, or is there anything you'd like to correct or add?

  7. Captain (synthetic)

    That's right. Just one thing: the go-around call was closer to five hundred feet than six hundred.

  8. Tracy (AI)

    Got it, about five hundred feet. Thank you so much for your time today, and for taking the trouble to file this report — what you have shared genuinely helps make things safer for everyone. A member of the safety team will follow up with you if anything more is needed. Take care, and goodbye.

Step 4 of 7

Listen in, or take the call over

Synthetic case · DEM 214 go-around · dispatcher's interview, live

While Tracy interviews the dispatcher who released the flight, an investigator can listen silently, see who is speaking and take the call over on their own phone. TRACE Ai saves both sides as text, line by line, while the call runs; the audio is not recorded. The dispatcher was involved but did not file the report, so Tracy's introduction says so instead of thanking them for a report.

On this screen
  1. Live: Tracy is interviewing
  2. Take the call over on your own phone
  3. Silent listening; the audio is not recorded
  4. Involved, not the reporter, and the introduction says so
TRACE Ai voice interview in progress with synthetic demonstration data, for a synthetic dispatcher. The Live call panel reads Tracy is interviewing, with Stop listening, Mute and Take over buttons, a green speaking light for the participant and the note that listening is silent and the audio is not recorded. Beside it are the five-question plan and the transcript so far, from the consent notice to the dispatcher's account of the release. Below, the Lifecycle panel lists the Approve the plan step with its button disabled, as TRACE Ai shows it after approval, and the call in progress and connected.
Listen in or take overDemonstration data · Click to enlarge
Step 5 of 7

Interviews in Spanish, French or Portuguese

Synthetic case · DEM 316 turbulence · cabin crew interview

An investigator can hold a TRACE Ai interview in Spanish, French or Portuguese, chosen when the interview is drafted. In this second synthetic case, a cabin crew member is interviewed in Spanish (Mexico) about a turbulence encounter, from the consent notice onwards, with an English machine translation under each line for investigators. The participant's own words remain the record.

On this screen
  1. Held in Spanish (Mexico), with a machine translation
  2. The English translation sits under each line
  3. Tracy introduces herself as an AI, in Spanish
TRACE Ai voice interview transcript with synthetic demonstration data, held in Spanish (Mexico). A bar reads English translation 11 of 11 lines (machine translation), with a ticked box to show it under each line. The transcript shows the Spanish consent notice, the synthetic cabin crew member consenting, Tracy's Spanish introduction, and the crew member's account of clearing the service and securing the carts after the seat-belt chime, about two minutes before the turbulence began, each with an English translation beneath. Beside it are the call-ended notice and the booking options.
Interviews in SpanishDemonstration data · Click to enlarge
Step 6 of 7

Cited cross-interview analysis, coded to HFACS and TEM

Synthetic case · DEM 316 turbulence · three interviews

When the interviews are finished and an investigator asks, TRACE Ai drafts a timeline, agreements, conflicts, gaps, candidate factors and recommendations across the accounts. Factors are coded against Threat and Error Management and HFACS, and every item cites the interview lines it rests on. Items without a supporting line, or with blame or hindsight language, are withheld, and nothing becomes a finding until an investigator accepts it.

On this screen
  1. A draft: TRACE Ai does not decide findings
  2. TEM environmental threat, accepted by the investigator
  3. Each item cites the interview lines it rests on
  4. HFACS crew resource management candidate, still to review
TRACE Ai Cross-interview analysis panel with synthetic demonstration data, drafted from three interviews (captain, first officer, cabin crew) with five of eighteen items reviewed. A note says the draft is for investigator review and TRACE Ai does not decide findings. Tabs cover timeline, agreements, conflicts, gaps, factors and recommendations. The Factors tab shows an accepted Threat and Error Management environmental threat and an HFACS crew resource management causal candidate, each with citation chips, the cited interview lines (one in Spanish) and Accept, Reject and Reword buttons.
Cited cross-interview analysisDemonstration data · Click to enlarge
Analysis excerpt (text)The Factors tab of the synthetic DEM 316 analysis, as on the screen above.

F1 · Convective build-ups on the arrivalThreat and Error Management · Environmental threatContributoryAccepted

Both pilots saw a line of build-ups on the radar along the arrival before the descent, and the crew asked for a deviation to the north. The captain says it took them around the worst of it, but not all of it.

  • I1 · Captain · turn 4"Before top of descent we could see build-ups on the radar along the arrival, so we asked for a deviation to the north. The deviation took us around the worst of it, but not all of it."
  • I2 · First Officer · turn 4"The radar showed a line of build-ups across the arrival before we started down. We briefed a deviation to the north and the captain asked for it."

F2 · No direct call to stop the serviceHFACS · Crew resource managementCausal candidateTo review

Step 7 of 7

A person signs off the review

Synthetic case · DEM 214 go-around

With every stage drafted, including a review pack for Review & follow up, TRACE Ai's autopilot leaves the review stage open. Its notes propose a review meeting, marked not agreed. The investigator decides the findings and completes the review stage, which the autopilot never does.

On this screen
  1. 7 of 8 stages completed; the review waits for a person
  2. Proposed, not agreed
  3. Only a person ticks this stage complete
  4. Ready for the investigator's sign-off
TRACE Ai investigation page with synthetic demonstration data for a fictional unstable approach and go-around of flight DEM 214 at Demo Field. The investigation guide shows 7 of 8 stages completed: Plan & prepare to Write report are ticked and tagged Autopilot draft, and Review & follow up is open and unticked, with working notes headed Autopilot draft that propose a review meeting, not yet agreed. Below, the Autopilot panel is marked Finished and reads: Every stage is drafted, and Review & follow up is ready for your sign-off.
A person signs off the reviewDemonstration data · Click to enlarge
Book a demo

Book a demo. Tracy will call you.

Hear the interviewer for yourself. Tracy will phone you, introduce herself as an AI, and walk you through a short interview about a fictional safety report, the way she speaks with a reporter after an occurrence. Afterwards, we show you the transcript and the investigation it lands in.

  1. 01

    Leave your number

    Tell us where to reach you.

  2. 02

    Agree a time

    We email you to agree a time. No account needed.

  3. 03

    Talk to Tracy

    A short interview about a fictional safety report, at your own pace.

  4. 04

    See the other side

    We walk you through the transcript and the investigation it is filed into.

Ask Tracy to call you

We email you to agree the time of your call.

Start with + and your country code. Tracy calls this number only.

Read how we use your details in the AI demo call privacy notice.

Controlled trial

Try TRACE Ai on your own past reports.

Run closed investigations through TRACE Ai side by side with your current process, then compare the findings, the actions and the time each took.

  1. 01

    Choose the reports

    Pick closed safety reports your team has already investigated, with the manuals that applied.

  2. 02

    Run both

    TRACE Ai builds each case from those reports and manuals while your investigators keep their usual process.

  3. 03

    Compare

    Review the root causes, actions and time taken together, and decide what to measure next.

Sign up for the trial

Read how we handle your information in our Privacy Policy.

An example workflow · TRACE Ai

From a safety report to a reviewable investigation.

A new occurrence needs an investigation. The investigator has a report, supporting documents, and people to hear from.

  1. 01

    Build the case

    Open the investigation, attach reports, logs, and photos, and start the autopilot to draft each stage and list the evidence still needed.

  2. 02

    Hear from the people involved

    Approve the people TRACE Ai proposes and the questions drafted for each. Tracy phones them at the time they choose, and each transcript is filed into the evidence.

  3. 03

    Prepare the next decision

    Compare the accounts with each other and with the evidence, then carry the investigation into interventions, the report, and review.

The resulting record

An investigation record with source evidence, consented interview transcripts, and a traceable basis for recommendations.

The human decision

The investigator assesses the evidence, approves the findings, and signs off the investigation.

Walk through this scenario with us ↗
How it works · TRACE Ai

A codified investigation, with the call built in.

Every investigation follows the same eight-stage human-factors method, and when an investigator starts the autopilot, it drafts each stage into the working notes. At Collect & debrief it proposes who to interview, books and calls each person once an investigator approves their plan, and stops at a review pack only a person can sign off.

  1. Plan & prepare
  2. Collect & debrief
  3. Analyse data
  4. Reconstruct context
  5. Test & validate
  6. Develop interventions
  7. Write report
  8. Review & follow up
Inside Collect & debrief
Before the call
  • Who to hear fromThe autopilot proposes up to six people whose first-hand accounts would close specific gaps, and says why each one is needed.
  • Questions from the recordEach person's questions are drafted from the investigation's own record, aimed at gaps only they can close, with blame and liability wording screened out.
  • Approved, then lockedNo one is contacted until an investigator approves that exact question set, which is checked again before dialling and when the call connects.
On the call
  • AI disclosure and consentThe consent notice plays in Tracy's voice when the call connects, and she says she is an AI. A clear no or a withdrawal ends the interview, and pressing 9 or asking for a person puts them through to the safety team or flags a follow-up.
  • Their own account firstTracy invites the person's full account first, then works through the approved questions conversationally, skipping what is already answered and asking open, non-leading follow-ups.
  • Read-back and correctionTracy reads back a summary of the account and invites the person to correct it before the call ends.
After the call
  • Filed as evidenceBoth sides are saved as text while the call runs, no audio is kept, and the transcript is filed into the investigation's evidence.
  • Originals keptIf an investigator corrects a line, the transcript is filed again with the correction beside the original wording.
  • Signed off by a personThe autopilot stops at a review pack, and an investigator reviews the evidence, decides the findings and signs off the review stage.
The eight stages of a TRACE Ai investigation, with the interview step of Collect & debrief opened up.
Capabilities

The details that keep work moving.

Each capability keeps the investigation, its evidence, and the people responsible for it in one record.

Run every investigation the same way

Each investigation follows an eight-stage human-factors method, from planning to review and follow-up, and every stage sets out its purpose, working outputs and a notes outline. Ask about the occurrence, crew, aircraft, evidence or prior findings, and the answers draw on the investigation record and the stage you are in. Revisit any stage as evidence changes; stages the autopilot drafted stay marked until an investigator reviews them.

Let the autopilot do the groundwork

Start the autopilot and TRACE Ai drafts each stage into the working notes. It proposes up to six people whose first-hand accounts would close specific gaps and explains why each is needed; once you add their contact details and approve their questions, it sends booking links, follows each call and files each transcript into the evidence. It then writes up the accounts and drafts the remaining stages, stopping at a review pack for an investigator to sign off.

Interview by phone with Tracy

Tracy phones each participant and holds a natural conversation built on the approved questions. She invites the person's own account first, asks open, non-leading follow-ups in their words, skips what they have already covered, and reads back a summary for them to correct. She can also interview in Spanish, French or Portuguese, with an English translation under each line for investigators.

Approve the questions, then let them choose a time

Each person's questions are drafted from the investigation record and the linked occurrence, aimed at the gaps only they can close, with questions about blame, liability, or discipline screened out. Nobody is contacted until an investigator approves the plan, and the approval is checked again before dialling. Participants then pick a call time from an emailed link without creating an account, Tracy calls at that time, and unanswered invitations get reminders.

Listen in, take over, file the record

Follow both sides of the call as text, or listen in live without the participant hearing you; the audio is never recorded. If a call needs a person, take it over: Tracy tells the participant she is connecting them and your phone rings. When the call ends, the transcript is filed into the evidence, corrections sit beside the original words, and investigators can download a restricted transcript record.

Compare the accounts, with every item cited

When you ask, TRACE Ai drafts one timeline from the finished interviews, where the accounts agree or differ, and the questions still open. It codes candidate factors against HFACS and Threat and Error Management, and drafts recommendations linked to them, each with a suggested owner and a measure of effectiveness. Every item cites the interview lines it rests on, and investigators accept, reject or reword each one.

Open every call with consent

Each call opens with a fixed consent notice in Tracy's voice, and the person is heard only after it has played. Their own words are kept as the record of their decision, and a clear no or a withdrawal ends the interview with a flag for an investigator. Pressing 9 or asking for a person puts them through to the safety team, or Tracy says the team will follow up.

Build the evidence context

Bring in reports, logs, PDFs, and photos. Images are described by a vision model so later answers can use what they show, and TRACE Ai keeps a live list of the evidence still outstanding.

Preserve accountability

Safety content is encrypted field by field inside the application, key actions go to a cryptographically chained audit ledger, and role-based access and single sign-on control who sees each record.

Applied AI

AI does the legwork. People decide.

TRACE Ai puts AI where investigations lose the most time: gathering accounts, organising evidence, and drafting each stage. Investigators approve the questions and every call, and own the conclusions.

A conversation, not a questionnaire

Tracy debriefs the way experienced investigators do: the person's own account first, open follow-ups in their words, questions about what they knew and noticed at the time rather than with hindsight, and a read-back so they can correct anything before she closes.

A structured investigation method

Each investigation follows an eight-stage human-factors method, from planning and debrief through analysis, context reconstruction, interventions, report, and review. The investigation assistant, the autopilot and the interview question drafts are all given the active stage's guidance and the investigator's notes.

Boundaries set in code

The consent notice, plan approval, call time limits, and ending the interview after a clear no at the consent question or a press of 9 are enforced by the application, not left to the model. In the cross-interview analysis, an item without a supporting participant line, or with blame, hindsight or credibility language, is withheld before an investigator sees it. If a voice provider fails, the call closes and the interview is flagged for follow-up.

Evaluate TRACE Ai

Make the next step measurable.

Agree a representative sample and a baseline from your current process. These are proposed evaluation measures, not published performance results.

Bring a representative example

A representative closed case with permission to use its evidence.

Inspect the complete workflow

Rebuild the evidence trail, hear Tracy interview a volunteer, and review the recommendations.

Compare with your baseline

Investigator preparation time; missing evidence; reviewer corrections.

Questions & answers

Before you book a demo.

Can Tracy call me for a demo?

Yes. Leave your number with the form on this page and we email you to agree a time. Tracy calls you only at that time, for a short interview about a fictional safety report.

Who is Tracy?

Tracy is TRACE Ai's voice interviewer. She phones the people an investigation needs to hear from, introduces herself as an AI, and holds a conversation built on the questions an investigator approved. Her transcripts are filed into the investigation's evidence.

Does Tracy read a fixed script?

No. Tracy words the approved questions conversationally, asks open follow-ups, and skips what the person has already covered. The consent notice is fixed text she cannot change. Teams that need every question read exactly as approved can switch to a scripted mode.

Does the AI decide investigation findings?

No. The autopilot drafts each stage and the cross-interview analysis proposes factors and recommendations, but each analysis item waits for an investigator to accept, reject or reword it. Investigators review the evidence, decide the conclusions, and sign off the review stage, which only a person can complete.

What controls apply to voice interviews?

An investigator approves each person's questions before contact, and the approval is checked again before dialling. Calls go only to authorised contacts and open with a fixed consent notice. A decline or withdrawal ends the interview with a flag for follow-up; pressing 9 or asking for a person puts them through to the safety team or flags it the same way. Investigators can listen in on a call in progress and take it over.

Does TRACE Ai require AURORA Ai?

No. TRACE Ai is a standalone safety application that owns its safety cases, investigations and interview records.

How do TRACE Ai and DELPHI Ai work together?

TRACE Ai runs with a DELPHI Ai service and needs it to start. It sends bounded investigation context to DELPHI Ai for cited manual retrieval, investigation drafting and drafts of the document changes a safety action calls for, then preserves the resulting answer and evidence in its own record. Tracy's calls, cross-interview analysis and transcript translation use separately configured voice and language models, and the analysis replaces participants' recorded names and contact details with labels before the model sees the text. Cases, evidence, investigation stages, and interview records belong to TRACE Ai.

Which investigation stages does TRACE Ai follow?

TRACE Ai structures an investigation as preparation, collection, analysis, reconstruction, testing, recommendations, reporting, and review. Investigators can return to an earlier stage as evidence changes, and stage notes and progress stay with the investigation record.

Which languages can Tracy interview in?

English, Spanish, French and Portuguese, in US and UK English, Mexican and European Spanish, Canadian and European French, and Brazilian and European Portuguese. The consent notice and closing are spoken in the interview's language, and Tracy follows a participant who switches to English. Investigators read an English machine translation under each transcript line and can hide it, while the original words remain the record.

What happens to TRACE Ai when our operation grows or changes?

As your fleet, routes, rules or teams change, we adapt TRACE Ai to match, and you approve each change before it goes live.

See TRACE Ai in action

Start with your workflow.

A focused walkthrough, with the people and decisions that matter to your team.

Request a demo