AI Engineer

Job Description

Role Summary

We are seeking an AI/ML Engineer with 3 to 6 years of experience to develop and manage the generative AI models that power our OCM platform. In this role, you will develop and maintain AI pipelines for content generation and evaluation, working with state-of-the-art ML tools. You will ensure our AI models produce high-quality, multilingual content and continuously improve through feedback loops.

Responsibilities

• Design and implement end-to-end generative AI pipelines for creating personalized marketing content. This includes data preprocessing, model inference, and post-processing of outputs for quality assurance.
• Leverage pre-trained models (e.g., via Azure OpenAI or Hugging Face models) and customize them for the domain needs. 
• Create mechanisms to automatically score or classify generated content for relevance, quality, and compliance. Implement evaluation workflows to filter out inappropriate or low-quality AI-generated content before it reaches end-users.
• Incorporate feedback from users and domain experts into the model improvement process. For example, implement prompt adjustments, model retraining, or rules based on what content performs best, to continuously refine output.
• Ensure the AI pipeline can generate content in multiple languages. Adapt models or use translation services as needed so content stays accurate and contextually appropriate across languages.
• Deploy models and AI services into production (using Azure services like Azure Functions or container deployments). Monitor model performance (latency, accuracy) and optimize as required. Troubleshoot and resolve issues in model inference in production environments.
• Work closely with the cross functional teams to develop AI capabilities into the application.
• Stay updated with the latest developments in AI/ML (new models, techniques, prompt engineering strategies) and experiment with improvements that could enhance our products intelligence and efficiency.
• Document model architectures, experiment results, and maintain version control for models and data (ensure we know which model version is in production). Follow best practices for model lifecycle management.

Qualifications

Required Skills

• 3–6 years in AI/ML engineering or applied data science, with hands-on experience in building and deploying machine learning models.
• Strong proficiency in Python and deep learning frameworks.
• Experience with generative models (Transformers, GPT-based models like GPT-Image-1, GPT-4o etc or similar) is essential.
• Familiarity with the Hugging Face ecosystem (libraries like Transformers, datasets). Experience with DSPy or other AI orchestration frameworks is a plus.
• Experience using cloud AI services, particularly Azure OpenAI Service and Azure Cognitive Services, to leverage large language models and other AI capabilities.
• Understanding of RAG and vector databases (e.g., pgvector, Pinecone, etc.) and how to use them for storing embeddings or enabling semantic search.
• Ability to build or utilize APIs (using Python FastAPI or Flask) to serve AI model functionality.
• Experience in prompt engineering – crafting and refining prompts to get better AI output – and fine-tuning models for improved performance.
• Understanding of active learning or feedback loop mechanisms to improve models over time.
• Prior experience deploying ML models in a production environment. Comfortable with best practices for scaling ML services, monitoring (logging, performance tracking), and troubleshooting live issues.
• Excellent analytical skills to diagnose issues with model output or data. Ability to experiment methodically and come up with creative solutions in an emerging field.

Good to Have

• Familiarity with content validation against guidelines or compliance requirements (e.g., understanding of medical, legal, regulatory checks for marketing content). Experience with LMRC or similar review processes is a plus.
• Some knowledge of data pipelines, ETL, or data preprocessing techniques, especially for unstructured data (Especially Image data).
• Interest or background in healthcare or marketing technology. Understanding the unique challenges of healthcare marketing (such as privacy or compliance) can help in building better AI solutions.
• Experience with ML lifecycle tools (MLflow, Kubeflow) or experiment tracking, and familiarity with DevOps for ML (CI/CD for model deployment)are nice to have.