About Me
Turning Machine Learning Into Scalable Products
👋 Hi, I’m Hichem Maiza, a Senior Machine Learning Engineer with a strong background in deploying scalable NLP and speech processing systems in production environments.
With over 8 years of experience in the industry, I specialize in machine learning optimization, model deployment, and system integration. My work spans across Natural Language Processing (NLP), Automatic Speech Recognition (ASR), and large language model (LLM) integration. I’m particularly passionate about turning state-of-the-art research into robust and efficient products.
What I Do
Currently at Oppscience, I Contribute to the optimization and deployment of ML models, focusing on:
Scaling ML pipelines to handle thousands of documents via batching and multithreading
Integrating LLMs like Qwen & LLama for text analysis and smart agent features.
- Developing microservices in Python and Java  for model APIs
Previously, at Enghouse Interactive, I developed ML services for sentiment analysis, topic classification, NER, and emotion detection, managing end-to-end production pipelines and contributing to the company’s internal ML toolkits.
My journey began at DeepOR where I worked on TTS, hotword detection, diarization, and drug NER. I also have research experience from my time at LITIS Lab, where I contributed to a published IEEE paper on obstacle detection using stereo vision and CNNs.