The ASR Engineer joins Omilia's Delivery ASR team to own the development, tuning, testing, deployment, and live performance of Omilia's ASR technology for enterprise customers' production IVR and voice-channel deployments. This role carries substantial responsibility and ownership: the ASR Engineer leads ASR solutions end-to-end, coordinating with other teams as needed, and is trusted to make sound judgment calls and manage their own work to meet tight deadlines.

Accountabilities

- Own end-to-end ASR tuning, model, and grammar delivery for a portfolio of live enterprise IVR/voice deployments, from requirements through go-live and ongoing optimization.
- Own root-cause investigation and resolution of production ASR issues, ensuring rapid recovery of customer-facing voice channels under time pressure.
- Own the technical relationship for assigned enterprise accounts, acting as the primary ASR delivery point of contact with customer and partner stakeholders.
- Drive measurable WER/SER improvements against production baselines through acoustic and language model design and fine-tuning, including modern neural approaches.
- Contribute to and help scale delivery automation to reduce manual turnaround across the team's release cadence.
- Ensure ASR solutions perform reliably across the team's production infrastructure and deployment environments spanning multiple global regions.

Key Responsibilities

- Tune ASR grammars, acoustic models, and language models for new locales, domains, and customer-specific vocabularies.
- Design, train, and fine-tune acoustic and language models, including modern neural approaches, with measurable WER/SER improvement.
- Investigate and resolve production ASR issues such as recognition failures and showstopper defects, engaging directly with internal and external customer/partner stakeholders.
- Run customer working sessions and translate technical tradeoffs for non-technical stakeholders.
- Represent ASR delivery in cross-functional forums with Product, Support, and Project Management.




- Build and maintain grammar-generation and other delivery automation pipelines to speed up release cycles.
- Monitor live ASR performance and proactively flag and address degradation before it affects customers.
- Deploy and validate ASR updates across production infrastructure spanning multiple integral regions.
- Document tuning decisions, model versions, and configuration changes for auditability and knowledge sharing.

Required:

- 5+ years of hands-on ASR/speech engineering experience, with substantial experience in production/live IVR or voice-channel deployments.
- Demonstrated track record supporting customers directly, ideally in regulated industries (banking, insurance, healthcare).
- Strong hands-on experience with ASR platforms such as Kaldi (or equivalent) — language modeling, acoustic modeling, grammar/lexicon design, and decoder tuning are required, not "nice to have."
- Solid programming ability (Python preferred; Bash, Java, or C++ also acceptable) and comfort in Linux environments.
- Academic background in Computer Science, Electrical Engineering, Computational Linguistics, Speech & Language Processing, or equivalent hands-on experience — a degree does not substitute for production experience at this level.
- Excellent spoken/written English and strong stakeholder communication skills; this role speaks directly with enterprise customers and partners.
- Comfortable owning ambiguous, production-critical problems independently, prioritizing across multiple accounts, and working under demanding deadlines.

Nice to have:

- Experience with modern neural ASR frameworks (e.g. wav2vec2, Whisper, or similar end-to-end models).
- Exposure to telephony/CCaaS platforms and IVR architectures.
- Familiarity with cloud infrastructure and MLOps practices for model deployment and monitoring.
- Additional spoken languages beyond English.

- Fixed compensation;
- Long-term employment with the working days vacation;
- Development in professional growth (courses, training, etc);
- Being part of successful cutting-edge technology products that are making a global impact in the service ind

📌 Automatic Speech Recognition Engineer (España)
🏢 Omilia
📍 España