Data Scientist - GEN AI

We are looking for an AI Engineer with experience in Speech-to-text and Text Generation to solve a Conversational AI challenge for our client based in EMEA. The focus of this project is to transcribe conversations and leverage generative AI-powered text analytics to drive better engagement strategies and decision-making. The ideal candidate will have deep expertise in Speech-to-Text (STT), Natural Language Processing (NLP), Large Language Models (LLMs), and Conversational AI systems. This role involves working on real-time transcription, intent analysis, sentiment analysis, summarization, and decision-support tools. The core responsibilities for the job include the following: Conversational AI and Call Transcription Development: Develop and fine-tune automatic speech recognition (ASR) models Implement language model fine-tuning for industry-specific language. Develop speaker diarization techniques to distinguish speakers in multi-speaker conversations. NLP and Generative AI Applications: Build summarization models to extract key insights from conversations. Implement Named Entity Recognition (NER) to identify key topics. Apply LLMs for conversation analytics and context-aware recommendations. Design custom RAG (Retrieval-Augmented Generation) pipelines to enrich call summaries with external knowledge. Sentiment Analysis and Decision Support: Develop sentiment and intent classification models. Create predictive models that suggest next-best actions based on call content, engagement levels, and historical data. AI Deployment and Scalability: Deploy AI models using tools like AWS, GCP, and Azure AI, ensuring scalability and real-time processing. Optimize inference pipelines using ONNX, TensorRT, or Triton for cost-effective model serving. Implement MLOps workflows to continuously improve model performance with new call data. Requirements: Strong experience in Speech-to-Text (ASR), NLP, and Conversational AI. Hands-on expertise with tools like Whisper, DeepSpeech, Kaldi, AWS Transcribe, and Google Speech-to-Text. Proficiency in Python, PyTorch, TensorFlow, and Hugging Face Transformers. Experience with LLM fine-tuning, RAG-based architectures, and LangChain. Hands-on experience with Vector Databases (FAISS, Pinecone, Weaviate, ChromaDB) for knowledge retrieval. Experience deploying AI models using Docker, Kubernetes, FastAPI, and Flask. Soft Skills: Ability to translate AI insights into business impact. Strong problem-solving skills and ability to work in a fast-paced AI-first environment. Excellent communication skills to collaborate with cross-functional teams, including data scientists, engineers, and client stakeholders.