AI/ML Intern

HafizApp is seeking a motivated Machine Learning Intern with a focus on speech and audio machine learning. This internship offers an opportunity to gain hands-on experience in training, evaluating, fine-tuning, and deploying machine learning models. Candidates with experience or interest in speech, audio, or language-related ML are strongly preferred. The position is available both remotely and on-site, with the on-site location based in Lahore, Pakistan. The internship is open to all qualified male and female candidates, offering a monthly stipend ranging from PKR 90,000 to PKR 180,000.

Key Responsibilities
- Train, fine-tune, test, and deploy machine learning models effectively.
- Handle large speech, audio, or language datasets, including preprocessing and conducting quality checks.
- Experiment with model parameters, hyperparameters, and tuning strategies to optimize performance.
- Evaluate models using appropriate metrics, compare experimental results, and perform detailed error analysis.
- Support ongoing model monitoring, performance improvements, and basic detection of data drift.
- Work with linguistic elements such as vocabulary, tokenization, pronunciation, phonemes, transcripts, and text normalization when applicable.
- Maintain version control of models, document experiment results, and ensure reproducibility of configurations.
- Collaborate closely with engineering, product, annotation, and quality assurance teams to ensure alignment and integration.

Required Qualifications
- Strong foundational knowledge of machine learning concepts, including training, validation, overfitting, and generalization.
- Practical experience in training or fine-tuning machine learning models.
- Understanding of ML evaluation metrics and the ability to conduct error analysis.
- Proficiency in Python and experience with ML frameworks such as PyTorch, TensorFlow, or similar.
- Experience in handling and preprocessing large datasets efficiently.
- Basic understanding of model deployment processes and monitoring techniques.

Preferred Qualifications and Benefits
- Experience with automatic speech recognition (ASR), speech-to-text systems, audio processing, pronunciation assessment, or related speech ML tasks.
- Familiarity with Transformer architectures, Hugging Face libraries, pretrained models, and embeddings.
- Knowledge of word error rate (WER), character error rate (CER), tokenization, phonemes, transcripts, and sequence modeling concepts.
- Experience using Git, GPUs or cloud platforms, containerization technologies, experiment tracking tools, or MLOps practices.
- Demonstrated practical experimentation through university projects, research, internships, Kaggle competitions, or ML portfolios.

Candidates should be currently pursuing or have recently completed a degree in Computer Science, Artificial Intelligence, Machine Learning, Data Science, Software Engineering, Electrical Engineering, or a related field. Strong analytical and problem-solving skills, along with a willingness to learn and experiment, are essential. The ability to document findings clearly and collaborate effectively with both technical and non-technical teams is highly valued.

This internship provides a valuable opportunity to develop practical skills in a dynamic environment focused on speech and audio machine learning. The role is based in Lahore for on-site work but also offers remote flexibility. The compensation ranges from PKR 90,000 to PKR 180,000 per month, reflecting the candidate’s experience and skills.