Aiphoria

About the Company

Aiphoria delivers an infinitely scalable workforce of AI-powered virtual employees for enterprise-grade Customer Support, Sales, Payments Management, Mass Recruitment, and other communication-intensive operations. Designed as voice-first workers, Aiphoria Pro agents can place phone calls, write messages, and respond to emails. They hold conversations with near-human quality while offering lower costs and virtually unlimited scalability — significantly boosting overall business efficiency. The Aiphoria Pro Platform also enables businesses to create their own agents using a zero-code design tool and a comprehensive set of analytical dashboards. Aiphoria’s globally distributed team includes some of the industry's most experienced AI and ML professionals, bringing over 15 years of hands-on expertise. Founded: 2022 Located: UK, UAE, Cyprus.

Listed Jobs

Company Name

Aiphoria

Job Title

Speech Data Engineer

Job Description

Job title: Speech Data Engineer Role Summary: Responsible for sourcing, evaluating, curating, and preparing multilingual speech datasets for ASR and TTS development, ensuring high-quality annotations and metadata, and supporting data pipelines and infrastructure. Expactations: - Deliver reliable, clean speech datasets that meet model training specifications. - Maintain versioned, accessible data repositories and documentation. - Collaborate with internal teams and external providers to optimize data quality and cost. Key Responsibilities: - Identify and assess unique data sources for speech data acquisition. - Collect, segment, diarize, and preprocess raw audio using tools such as VAD, Pyannote, whisper, etc. - Run speech recognition and pseudo‑labeling; work with crowdsourcing platforms to refine annotations. - Ensure transcription accuracy, speaker diversity, and consistent metadata across languages. - Align dataset specifications with ASR/TTS model needs and support data infrastructure (e.g., DVC). - Evaluate external datasets, support make‑vs‑buy decisions, and maintain clear records of datasets’ provenance. - Contribute to dataset versioning, organization, and lifecycle management. Required Skills: - Hands‑on experience with speech segmentation, diarization, and labeling tools (VAD, Pyannote, Whisper). - Proficiency in acoustic quality metrics: SNR, spectral analysis, and other signal‑to‑noise indicators. - Strong understanding of differences between ASR (noisy) and TTS (clean) data requirements. - Familiarity with data versioning (e.g., DVC) and pipeline maintenance. - Ability to work with multilingual audio recordings and diverse speaker populations. - Excellent documentation and version control practices. Required Education & Certifications: - Bachelor’s or Master’s degree in Computer Science, Speech Engineering, Linguistics, or related field. - Certifications in speech processing, data science, or related domains preferred (e.g., Microsoft Certified: Data Analyst Associate, TensorFlow for Cloud & Edge).

London, United kingdom

Remote

12-01-2026

Company Name

Aiphoria

Job Title

Machine Learning Engineer TTS

Job Description

**Job title**: Machine Learning Engineer – Text-to-Speech (TTS) **Role Summary** Develop and optimize advanced TTS models (FastPitch, FastSpeech 2, VITS, Glow‑TTS) to deliver natural, expressive voice output for a voice‑assistant platform. Collaborate with product, engineering, and data teams to integrate TTS technology, construct audio data pipelines, and continuously improve model performance across accents, dialects, and prosodic styles. **Expectations** - Design, train, and refine end‑to‑end TTS architectures using PyTorch. - Ensure models meet quality benchmarks (MOS, A/B testing) and operational standards. - Stay current with emerging TTS research and integrate innovative techniques into production. - Deliver documented code, performance reports, and reproducible pipelines. **Key Responsibilities** - Architect and implement TTS models (FastPitch, FastSpeech 2, VITS, Glow‑TTS). - Build and maintain robust audio data pipelines from recording to model training. - Collaborate with product managers for feature requirements and user‑centric design. - Conduct rigorous evaluation (MOS, A/B) and iterate models for accent, dialect, and prosody control. - Integrate and benchmark vocoders (Vocos, HiFi‑GAN, Mimi). - Publish research‑grade experiments and internal white‑papers. **Required Skills** - Python programming and PyTorch deep‑learning framework. - Expertise in TTS synthesis techniques and fast attention‑based models. - Proficiency in prosody control, rhythm, and emotional tone modeling. - Knowledge of normalization methods (FSTs, NN normalization). - Experience with TTS evaluation metrics (MOS, A/B testing). - Familiarity with vocoders and signal‑processing fundamentals. - Strong statistical modeling, language‑structure understanding, and voice‑data pipeline design. **Required Education & Certifications** - Bachelor’s or Master’s degree in Computer Science, Electrical Engineering, Machine Learning, or related field. - Preference for additional coursework or certifications in Speech Processing, Deep Learning, or Signal Processing. Related professional certifications (e.g., Deep Learning Specialization, ML Engineer Certification) are advantageous but not mandatory.

London, United kingdom

Remote

12-01-2026