Artificial Intelligence Engineer

Key Responsibilities

• Design, develop, and deploy scalable AI/ML models and algorithms for production.
• Work on end-to-end machine learning pipelines from data preprocessing to model evaluation and deployment.
• Implement and fine-tune LLMs and Indic TTS models
• Collaborate with cross-functional teams to integrate AI capabilities into existing products and workflows.
• Optimize models for performance, scalability, and accuracy.
• Work on optimizing prompts and implementing best practices for best outputs and reduce hallucinations and token costs
• Monitor and maintain deployed models, ensuring data drift and performance issues are handled proactively.

Required Skills and Qualifications

• Bachelor's or Master's degree in Computer Science, Data Science, or a related field from Tier 1 colleges like NIT, IIT or BITS.
• 2+ years of experience in developing and deploying machine learning or deep learning solutions.
• Strong programming skills in Python and proficiency with AI/ML frameworks like TensorFlow, PyTorch.
• Hands-on experience with data processing libraries (NumPy, Pandas).
• Should have worked with LLMs (specifically audio based) and TTS models. You should have knowledge about the best practices around prompting and fine tuning LLMs for most optimized results
• Knowledge of API development (FastAPI/Flask) for serving models in production.

Good to Have

• Past experience with fine tuning and deploying fast Indic TTS models is a huge plus.
• Solid understanding of MLOps practices, version control (Git), and CI/CD pipelines.
• Experience with Vector Databases, LangChain.
• Familiarity with containerization (Docker, Kubernetes) and model monitoring tools.
• Experience with cloud platforms such as AWS, GCP, or Azure for model training and deployment.
• Publications, Kaggle profile, or contributions to open-source AI projects.

This is a role in an very fast-paced and result oriented environment. You will be responsible for all of our AI initiatives around speech, analysis and learning. You would be building for potentially 10-50 million bharat users and the work will be focused around building contextual memory and great user experience. It will be a challenging but rewarding experience. We have access to H100 and T4 GPUs for training, deployment and fine tuning and you will be given a free hand in terms of implementing best practices