Reliable language data begins with rigorous linguistic work.
At Umbaji, voice AI starts with sound linguistic research carried out alongside the communities who actually speak the languages. We are seeking senior linguists who can connect scientific methods, trustworthy datasets and technology that serves real needs.
2026 RECRUITMENT | SENIOR LEVEL | CONTRACT ENGAGEMENT | APPLY BY 30 NOVEMBER 2026
Build the scientific foundations of African voice technology.
Umbaji develops automatic speech recognition, machine translation and speech synthesis technologies for West African languages. Through Yodi, our community-driven data platform and our research and evaluation work, we connect field data collection, native-speaker validation, scientific research and technical delivery. Senior linguists provide the expertise needed to turn these contributions into documented, consistent and usable linguistic resources.
Yodi. Speech recognition, translation and text-to-speech models and services built for practical applications.
Data Hub. Voice and text data gathered with language communities and reviewed before being used.
LomBench & research. Methods, datasets and evaluations that make language-system quality measurable.
The role
Act as a scientific and operational reference for assigned languages. Depending on project needs, you will help define methods, oversee corpus work and evaluate outputs produced by data and AI teams.
What you will contribute
- Investigate language structure : Analyse relevant phonetic, phonological, morphological, syntactic and semantic patterns while respecting natural variation.
- Build and document corpora : Define collection units, metadata, transcription conventions and criteria for meaningful language coverage.
- Lead transcription and annotation : Create usable guidelines, resolve ambiguous cases and review segmentation and annotation accuracy.
- Strengthen translation quality : Produce or review context-sensitive translations that preserve register, meaning, terminology and cultural nuance.
- Evaluate speech and language models : Classify ASR, TTS and translation errors, document their underlying causes and identify priorities for improvement.
- Maintain scientific quality : Establish peer review, inter-annotator agreement checks, decision logs and clear progress reporting.
- Work across disciplines : Collaborate with speakers, annotators, researchers and engineers to turn linguistic decisions into production-ready practices.
Target languages
Strong proficiency in at least one of the following languages is required. Please specify the varieties you know and your ability to understand, speak, read and write, wherever those skills are applicable to the language.
Adangbe, Adele, Akan, Akaselem, Anii, Lebanese Arabic, Baatonum, Bago-kusuntu, Dagbani, Delo, Ga, Gbe (Maxi), Gbe (Xwla), Ginyanga, Hausa, Igo, Kebou, Krache, Kusaal, Lama, Lelemi, Lukpa, Mampruli, Pidgin, Waama, Woudou, Yoruba
Who we are looking for
- Education. A minimum of a Master’s degree in linguistics, language sciences, translation, NLP or a closely related field. PhD holders and research backgrounds are particularly welcome.
- Expertise. Advanced command of at least one target language and proven experience in language analysis, documentation, transcription, translation or corpus work.
- Methodology. Ability to design protocols, defend scientific decisions and deliver rigorous, reproducible and well-documented work.
- Collaboration. Ability to work with native speakers and multidisciplinary teams. Good command of French; English is an asset.
- Additional strengths. Experience in speech technologies, phonetics, annotation, NLP, scholarly publishing or leadership of linguistic teams.
Working arrangement
This is a contractual engagement within Umbaji’s language team. The scope, deliverables and working arrangements will be determined by project requirements and the selected candidate’s profile. Onboarding on tools, guidelines and quality assurance procedures will be provided as needed.
How to apply
To help us assess your expertise, prepare:
- An up-to-date CV detailing relevant qualifications, experience, projects and research.
- A cover letter describing how your expertise can contribute to language research and speech technology.
- A language proficiency statement listing languages, varieties and speaking, listening, reading and writing proficiency.
- Where available, publications, corpora, annotation guides, professional references or samples of relevant work.
Application deadline: 30 November 2026 Consult Umbaji’s Careers page for application instructions once this position is published. Umbaji Careers
Bring your linguistic expertise to the development of AI that understands languages as they are truly spoken.
Recruitment notice | Language Research | Umbaji