evidoria

← Back to browse

Good practice Imported

Audimus — AI Speech Recognition Automates Angola's National Assembly Transcripts

Angola · Luanda · See the Angola profile · See the Luanda profile

Evidence: Descriptive / self-reported Top 94% 27/100 · Ask Evidence Copilot about this practice

Since 2017, Angola's National Assembly has used Audimus, an AI speech engine from Portugal's VoiceInteraction, to auto-transcribe plenary audio into the official record. Two independent outlets confirm the deployment, but no timing, accuracy or cost figures are public.

Audimus — AI Speech Recognition Automates Angola's National Assembly Transcripts

Details

Maturity
Established
Promoter
Assembleia Nacional de Angola (National Assembly of Angola)
Period
2017–ongoing
Keywords
parliamentary administration, public sector productivity, speech-to-text

Context

Angola's National Assembly (Assembleia Nacional) has deployed Audimus, an artificial-intelligence speech-recognition engine built by the Portuguese company VoiceInteraction, to automatically convert audio from plenary and committee sessions into written text, reportedly in use since 2017.

Objectives

Feed automated transcription directly into the production of official session minutes ('atas'), a task previously done through manual transcription, making legislative record-keeping faster and less dependent on manual note-taking.

Activities

The National Assembly has paired the tool with staff AI training as part of a wider modernisation push, alongside newer AI initiatives at Angola's tax authority and electricity utility.

Results

Angolan press describes the system as making record-keeping faster, but no source consulted publishes transcription accuracy, time saved, cost, or usage-volume figures for Audimus, and no public technical or governance documentation was found.

Conclusions

Audimus is one of Angola's longest-running named AI deployments in the public sector, but its outcomes remain undocumented; separately, the President of the National Assembly has cautioned that Angola's broader push into digital and AI tools carries risks to fundamental rights needing legislative safeguards — a general caution included here for honesty, though not stated as being about Audimus specifically.

Implementation
Indicative cost
Low (< €50k) — No public cost figures disclosed; use of an off-the-shelf commercial ASR product suggests lower relative cost than a bespoke build — conservative low estimate pending curator review.
Time to results
Long (> 3 years) — Reportedly in continuous use since 2017 through at least 2026 coverage — a long-running deployment.
Staffing & skills
Assembleia Nacional de Angola (National Assembly of Angola), VoiceInteraction (Portuguese ASR vendor)

Conditions for success

  • Use of an established commercial speech-recognition product (Audimus) rather than a bespoke build
  • Paired with staff AI training as part of a wider modernisation push

Common failure modes

  • No public accuracy, time-saved, cost or usage-volume figures for the deployment
  • No public technical or governance documentation found
  • National Assembly leadership has separately cautioned that Angola's broader AI/digital push carries fundamental-rights risks needing legislative safeguards (a general caution, not stated as being about Audimus specifically)

Where it fits

Governance type
national legislature
Scale
single institution (National Assembly)

Commonly funded by

National / regional programmes

Indicative funding routes for practices of this type — always check each programme's current calls and eligibility rules.

Do you run this practice? Claim it — verified implementers get a public contact pathway and can propose corrections.

Data sources

Where this practice's information was retrieved from, and when.

Attachments

Similar practices you may find useful