As one of the first members of JT AI Labs, you will design and build the data foundations of our audio collection activity, from the raw recording to the dataset delivered to the client:● Own audio storage on AWS: bucket design, encryption, versioning, lifecycle and retention rules, and reconciliation between the database and storage.● Build audio processing jobs: track alignment between speakers, resampling to client specifications, and format validation, transform quality control ad-hoc scripts into full blown quality control pipelines.● Design and run the dataset delivery pipeline: packaging to client spec, file naming, metadata manifests, pseudonymised speaker IDs, consent records per file, checksums, and secure transfer.● Guarantee traceability: every file can be traced to its speaker, project and consent, and consent withdrawal deletes every version of the affected data and logs the deletion.● Make sure all storage and processing complies with GDPR and EU AI Act data requirements.● Analyse datasets, spot quality problems early, propose solutions and iterate quickly.● Under the supervision of the Head of AI, work closely with the full-stack engineer, the delivery lead and clients to meet our goals on acceptance rate.● Contribute to the team’s best practices in testing, CI/CD, data versioning and deployment automation.
As one of the first members of JT AI Labs, you will design and build the data foundations of our audio collection activity, from the raw recording to the dataset delivered to the client:● Own audio storage on AWS: bucket design, encryption, versioning, lifecycle and retention rules, and reconciliation between the database and storage.● Build audio processing jobs: track alignment between speakers, resampling to client specifications, and format validation, transform quality control ad-hoc scripts into full blown quality control pipelines.● Design and run the dataset delivery pipeline: packaging to client spec, file naming, metadata manifests, pseudonymised speaker IDs, consent records per file, checksums, and secure transfer.● Guarantee traceability: every file can be traced to its speaker, project and consent, and consent withdrawal deletes every version of the affected data and logs the deletion.● Make sure all storage and processing complies with GDPR and EU AI Act data requirements.● Analyse datasets, spot quality problems early, propose solutions and iterate quickly.● Under the supervision of the Head of AI, work closely with the full-stack engineer, the delivery lead and clients to meet our goals on acceptance rate.● Contribute to the team’s best practices in testing, CI/CD, data versioning and deployment automation.
● You hold a Master’s degree (or equivalent) in Computer Science, Data Engineering, Applied Mathematics or a related quantitative field.● You have 3+ years of experience building and running data pipelines in production.● You are proficient in Python and comfortable with Num
Py and the data tooling around it.● You have hands-on experience processing audio or other signal data: ffmpeg, formats and codecs, sample rates, and signal quality metrics.● You have strong experience with AWS (S3, IAM, KMS, lifecycle policies) and infrastructure as code (Terraform or similar).● You are rigorous about data quality, reproducibility and traceability: checksums, manifests, versioning, idempotent jobs.● You are familiar with data protection requirements (GDPR, PII handling, pseudonymisation), and you are aware of the EU AI Act or keen to learn about it.● You have a pragmatic, analytical, and results-oriented mindset: you favor frequent iterative deliveries over long research tunnels.● You thrive in a startup setting: you enjoy building from scratch, taking ownership, and navigating ambiguity.● You enjoy working in cross-functional teams (Operations, Tech, Product) and can translate client specifications into reliable pipelines.● You are autonomous, curious, and comfortable in a fast-evolving ecosystem.● Professional proficiency in English, bilingual in French.
Nice to have● Experience with speech datasets for ASR or TTS: voice activity detection (Silero VAD or similar), speaker diarisation, and transcription tools.● Experience delivering datasets to AI labs or data buyers, and with their delivery conventions.● Experience with annotation tools (e.g. Label Studio) and human-in-the-loop quality workflows.● Experience with health-data or other regulated data environments.
Vous serez redirigé vers l'annonce d'origine.
Déposez votre CV : votre score face à cette offre, ce qui manque, puis une version adaptée au poste en un clic.
Analyser mon CV253 offres de Data Engineer à Paris
Les 253 offres de Data Engineer à Paris