The Senior Data Scientist builds production-grade ML and multimodal systems that support media workflows such as automated metadata extraction, video understanding, speech transcription, moderation, and recommendation. This role owns model development, MLOps integration, and production deployment while mentoring junior engineers. The engineer develops robust pipelines spanning NLP, CV, ASR, OCR, and video analytics, and integrates AI into CMS platforms, OTT apps, and media asset management systems.

Key Responsibilities

Build and deploy multimodal ML models spanning NLP, CV, ASR, OCR, and video analytics.
Develop robust pipelines for video frame extraction, shot detection, speech-to-text, speaker diarization, image tagging, content moderation, and multimodal embeddings (CLIP, SigLIP, VideoCLIP).
Implement RAG (Retrieval Augmented Generation) with multimodal indexing.
Optimize model latency for real-time content tagging or streaming workflows.
Implement batching, quantization, distillation, or GPU optimizations.
Build CI/CD pipelines for ML using GitHub Actions, Jenkins, or Azure DevOps.
Deploy models as microservices using Docker, Kubernetes, KServe, or FastAPI.
Integrate observability tools (Prometheus, Grafana) for model monitoring.
Integrate AI into CMS platforms, OTT apps, and media asset management (MAM) systems.
Create APIs for search, recommendations, auto-captioning, and content summaries.
Collaborate with product, content, and media engineering teams.
Mentor associate engineers and review code.

Must Have Skills

Python
PyTorch or TensorFlow
NLP and Computer Vision
Multimodal LLM Frameworks (LangChain, LlamaIndex)
Vector Search (FAISS, Pinecone, Weaviate)
Kubernetes and Docker

Good To Have Skills

Speech Recognition (ASR)
OCR
RAG (Retrieval Augmented Generation)
FastAPI
CI/CD Tools (GitHub Actions, Jenkins, Azure DevOps)
Observability Tools (Prometheus, Grafana)