Globiva

AI Engineer (RAG/NLP) - On-site in Gurugram

Globiva is seeking a skilled AI Engineer to join our team in Gurugram. This role focuses on the design, development, and deployment of cutting-edge Generative AI and Natural Language Processing (NLP) solutions. You will be instrumental in building advanced Retrieval-Augmented Generation (RAG) pipelines, automating document processing with OCR/ASR technologies, and creating intelligent knowledge-assist systems. We are looking for a candidate with robust practical experience in Python, Hugging Face Transformers, vector databases, and API deployment. The ideal engineer will manage the complete AI model lifecycle, from initial data preparation to ongoing production monitoring.

Key Responsibilities

• Design, construct, and refine RAG pipelines, encompassing prompt engineering, text chunking, retrieval mechanisms, reranking strategies, and rigorous evaluation using vector databases like Chroma DB or Qdrant, and Transformer-based LLMs including Llama, Mistral, or BERT family models.
• Engineer and deploy production-ready Automatic Speech Recognition (ASR) systems, such as Whisper-large-v3, for demanding call centre and voice-interaction use cases, prioritizing enhancements in both accuracy and response latency.
• Develop sophisticated Optical Character Recognition (OCR) and document digitization workflows, utilizing libraries like OpenCV and Tesseract, complemented by CNN/LSTM-based post-processing for efficient handling of unstructured PDFs and images.
• Build and deploy robust APIs using Fast API or Flask, ensuring seamless integration with existing systems and data sources like MongoDB.
• Orchestrate complex data and model workflows through Airflow, automating ETL processes, establishing evaluation pipelines, and managing periodic model retraining.
• Implement comprehensive CI/CD pipelines for the streamlined release of models and APIs, upholding stringent standards for testing, logging, and observability.
• Conduct thorough offline and online evaluations to assess critical metrics such as latency, accuracy, F1 score, ASR Word Error Rate (WER), and retrieval precision/recall, delivering detailed analytical reports with actionable recommendations.
• Collaborate effectively with product and operations teams to translate intricate business challenges into quantifiable ML objectives and well-defined Service-Level Agreements (SLAs).

Mandatory Skills

• Expertise in production-grade Python, PyTorch, and Hugging Face Transformers.
• Deep understanding of RAG principles, including text chunking, embedding strategy, retrieval techniques (dense and sparse), reranking, and evaluation frameworks.
• Hands-on experience with vector search technologies and data stores like Chroma DB or similar, coupled with strong data modeling and indexing skills.
• Practical experience in API development with Fast API or Flask, adhering to RESTful best practices, including authentication, rate limiting, and pagination.
• Proficiency in MLOps practices, utilizing Docker, CI/CD, Linux, Git, and tools for logging and monitoring AI services.
• Exposure to OCR and ASR systems, specifically with tools like OpenCV, Tesseract, and Whisper or comparable frameworks.
• Solid foundation in classical NLP and ML techniques, including tokenization, LSTMs/CNNs, XGBoost, and metric-driven approaches.

Preferred Skills

• Experience in fine-tuning large language models or encoders for tasks such as classification, summarization, and domain adaptation.
• Knowledge of prompt engineering, tool integration, and evaluation methodologies for LLM-powered applications.
• Familiarity with scaling retrieval systems to support low-latency, high-availability production environments.
• Experience in document question answering, email or call centre analytics, or enterprise knowledge management solutions.
• Understanding of agentic AI frameworks like Lang Chain, Lang Graph, or Crew AI.

Qualifications

• 2 to 5 years of hands-on experience in AI, ML, or NLP engineering, with demonstrated experience in production ownership.
• Bachelor's degree in Computer Science, a related technical field, or equivalent practical experience.
• A strong portfolio or GitHub profile showcasing shipped APIs, implemented models, and well-documented repositories with comprehensive test coverage.