4423281-Manager

Job Title: Manager – Agentic AI & Applied NLP Experience: 7–10 years Location: [Insert Location] | Type: Full-time Role Overview We are looking for a hands-on AI Manager with strong experience in NLP and GenAI to build intelligent, autonomous systems using LLM-powered agents. This role is ideal for someone with a builder mindset, who can lead the development of task-driven agentic workflows and coordinate across teams to deliver scalable AI solutions. Key Responsibilities Design and build multi-agent systems for real-world tasks such as document understanding, claims triage, compliance monitoring, and knowledge retrieval. Implement LLM-based agents that can plan, reason, interact with tools, and collaborate with other agents or humans. Develop core agent capabilities like tool calling, context retention, memory integration, and decision-making. Work with libraries like LangGraph, LlamaIndex, Semantic Kernel, or AutoGen to orchestrate and evaluate agent pipelines. Integrate agents with retrieval systems (RAG) and external APIs for tool-augmented execution. Collaborate with AI platform and backend teams to deploy and monitor these agents in production environments (Cloud Run, K8s, Vertex AI). Mentor junior team members and lead proof-of-concept efforts for agentic use cases. Contribute to reusable templates, design patterns, and architectural blueprints for scalable agentic systems. Qualifications 7–10 years of experience in AI/ML with deep focus on NLP, GenAI, or cognitive architectures. Proven experience with LLMs (OpenAI, Claude, Gemini, or open-source models) and prompt engineering. Hands-on experience with agent orchestration libraries (LangGraph, AutoGen, CrewAI, etc.). Understanding of tool use in LLMs (function calling, plugins, API interaction). Familiarity with cloud-native AI platforms (Vertex AI, SageMaker, Azure ML). Experience in integrating agents with vector databases (Pinecone, FAISS, Weaviate) and retrieval mechanisms. Comfortable translating fuzzy business problems into autonomous task pipelines.