(2025-Current) - Intern @Télecom Paris

• High-quality and expressive music generation on audio informed with structure.
• Dataset processing.

(2023-2025) - Researcher @CPqD

• Improved a hybrid ASR by adapting its LM on synthetic domain-specific text generated by LoRA of LLMs.
• Evaluated fairness of the company’s ASR on multi-accented speech in the Brazilian Portuguese language.
• Developed an accurate and efficient two-stage SSL-based speech emotion recognition system.
• Enriched the company’s call center customer profiler by developing a SOTA speech age and gender classifier.

(2021-2023) - Fellow Master @CPqD

• Implemented neural customer-oriented expressive TTS models for the Brazilian Portuguese language.
• Enabled customers to edit synthesized audios with character-level prosody control on the ONNX FastPitch.
• Conducted perceptual experiments to evaluate speech naturalness, emotion intensity, and speaker similarity.

(2020-2021) - Internship @LPS

• Enhanced lab automation by designing neural speech commands recognition systems.
• Encapsulated the command recognition system in a local private LoRa network for IoT applications.
• Enabled long distance voice control by developing a wearable prototype with an embedded microphone.

Skills

• Deep Learning: PyTorch, Tensorflow, Sci-Kit, HuggingFace, ONNX, Gradio
• Frameworks: Lightning, Amphion, Coqui, SpeechBrain, ESPNET, Kaldi
• Tools: Docker, Git, Cloud (AWS, GCP)
• Programming: Python, C, C++, MATLAB, LaTeX
• Languages: Portuguese, English, French
• Competences: Paper Implementation, Experiment Design, Team Collaboration

Education

(2021-2024) M.Sc. Computer Engineering @UNICAMP

(2019-2020) Excellence Scolarship Exchange Student @Télécom Paris

(2017-2021) B.Sc. Electronic Engineering @UFPB