Key Responsibilities

   LLM Model Optimization

• Lead fine-tuning and optimization of large language models (LLMs) to improve accuracy and robustness in call-center scenarios.
• Design and implement conversation pipelines and plugin architectures using LangChain and LangGraph for efficient, scalable query and retrieval workflows.

   Retrieval-Augmented Generation (RAG) & Knowledge-Based Construction

• Architect and build RAG workflows: set up vector stores, retrieval modules, and integrate them with LLMs to generate context-aware responses.
• Collect, clean, annotate, and preprocess datasets, design automated pipelines for ingestion and continuous knowledge base updates.

   System Integration & Engineering

• Develop production-grade Python services on Linux, writing clean, maintainable, and scalable code.
• Containerize and orchestrate model services using Docker and Kubernetes; establish CI/CD pipelines for automated builds, tests, and deployments.

   Performance Monitoring & Continuous Iteration

• Integrate and optimize third-party speech and language services, developing custom connectors as needed.
• Implement monitoring and logging systems; analyze key metrics (latency, throughput, accuracy) and recommend performance improvements.

Qualifications

   Experience

• Minimum 5 years of experience in software development, with at least 2 years in the AI domain.
• Programming & Frameworks
• Expert in Python and at least one deep learning framework (TensorFlow, PyTorch, or JAX).
• Hands-on experience with LLM fine-tuning, prompt engineering, LangChain, and LangGraph.

   Data & Algorithms

• Strong understanding of vector embedding techniques and experience with tools such as Qdrant, FAISS, Annoy, or Pinecone.
• Proven ability to design and maintain end-to-end data pipelines (collection, cleaning, annotation, splitting).

    Systems & Deployment

• Minimum 1 year of production-level software development on Linux.
• Proficient in Docker and Kubernetes, with experience designing microservices architectures.
• Familiarity with CI/CD platforms (Jenkins, GitLab CI/CD, GitHub Actions) is a plus.

Preferred Qualifications

• Deep knowledge of vector database internals and operations (e.g., Qdrant, Milvus, Weaviate, Pinecone).
• Experience deploying multi-tenant online services with auto-scaling.
• Familiarity with an additional backend language (Go, C++) or frontend frameworks (React, Vue).
• Background in integrating AI into financial, e-commerce, or telecom call-center platforms.
• Publications in related fields or significant open-source contributions.