Need Senior AI System Engineer ( Python, MCP, Langchain)
Visionsoft International
McLean, VA (In Person)
Full-Time
Skill Insights
Compare your current skills to what this opportunity needs—we'll show you what you already have and what could strengthen your application.
Job Description
Need Sr Python Developer with strong AI, MCP exp McLean, VA Hybrid Python Fundamentals (Must Have) Deep expertise in Python 3.10+, including asyncio, multithreading/multiprocessing, decorators, generators, and metaclasses Proficiency with foundational packages: NumPy, Pandas, Pydantic, httpx/requests, dataclasses, typing Strong grasp of clean code principles, SOLID design, and Pythonic idioms Experience writing unit/integration tests with pytest and maintaining high code coverage Familiarity with linting/formatting toolchains (ruff, black, isort, mypy) and pre-commit hooks Experience with dependency and environment management (Poetry, uv, pip, venv, conda) Agentic AI, LangChain & MCP (Core Focus) Proven hands-on experience with Model Context Protocol (MCP) designing, building, and maintaining MCP servers and clients Strong working experience with FastMCP for building Python-based MCP servers with tools, resources, and prompts Expert-level experience with LangChain (chains, agents, memory, retrievers, output parsers, LCEL) Experience with LangGraph for stateful, multi-agent, and graph-based agentic workflows Understanding of tool/function calling, structured outputs, and agent-to-agent communication patterns Experience integrating multiple LLM providers (Anthropic Claude, OpenAI, Azure OpenAI, Gemini, open-source models) Knowledge of RAG architecture: chunking strategies, embeddings, hybrid search, re-ranking, and evaluation ML/AI Foundations Working knowledge of machine learning fundamentals: embeddings, similarity metrics, classification, evaluation Familiarity with PyTorch, TensorFlow, or scikit-learn for model training/inference where needed Experience with Hugging Face ecosystem (Transformers, datasets, model hub) Understanding of prompt engineering, few-shot learning, and LLM evaluation frameworks (RAGAS, DeepEval, LangSmith evals) Cloud, DevOps & MLOps 4+ years deploying applications on AWS, Azure, or Google Cloud Platform (Lambda, ECS/EKS, Cloud Run, Azure Functions) Proficiency with Docker; working knowledge of Kubernetes and Helm CI/CD experience with GitHub Actions, GitLab CI, or Azure DevOps Experience with LLM observability and tracing tools (LangSmith, Langfuse, Arize Phoenix, OpenTelemetry) Familiarity with secrets management, rate limiting, and cost monitoring for LLM workloads Security & Responsible AI Experience implementing guardrails, input/output validation, and PII handling in AI pipelines Awareness of prompt injection risks and mitigation strategies in agentic/MCP-based systems Understanding of compliance considerations (SOC 2, GDPR, HIPAA) when handling sensitive data Collaboration & Leadership Experience mentoring engineers, conducting code reviews, and setting technical standards Ability to translate business problems into AI solution architectures Excellent communication skills with both technical and non-technical stakeholders Comfortable in Agile/Scrum delivery models with tools like Jira and Confluence Backend & API Development 5+ years building production APIs with FastAPI, Flask, or Django REST Framework Experience with streaming responses (SSE/WebSockets) for real-time LLM output Solid understanding of authentication/authorization mechanisms (OAuth2, JWT, API key management) Experience designing scalable microservices and event-driven architectures (Kafka, RabbitMQ, Celery) Data & Storage Strong SQL skills (PostgreSQL, MySQL) and experience with ORMs (SQLAlchemy) Hands-on experience with vector databases: Chroma, Pinecone, Qdrant, Weaviate, pgvector, or FAISS Experience with caching layers (Redis) and NoSQL stores (MongoDB, DynamoDB) Data preprocessing, ETL pipeline development, and working with structured/unstructured data Nice to Have Contributions to open-source AI/LLM projects (LangChain, MCP servers, etc.) Experience with fine-tuning (LoRA/QLoRA) or self-hosted model serving (vLLM, Ollama, TGI) Knowledge of A2A protocols, CrewAI, AutoGen, or other multi-agent frameworks Experience building Slack/Teams bots or IDE integrations powered by MCP Education & Experience Minimum 7-10 years of overall software engineering experience with strong Python expertise 3+ years of hands-on experience building LLM-powered or AI/ML applications in production Bachelor's/Master's degree in Computer Science, Engineering, AI/ML, or equivalent industry experience Demonstrated experience owning end-to-end delivery of AI products from design to deployment