Predictive & Agentic AI Engineer

Bright MLS, Inc•North Bethesda, MD

13h•$155,000 - $175,000

About The Position

Are you an accomplished AI practitioner with a passion for developing cutting-edge analytics, predictive machine learning, and multi-agent generative AI solutions? If you have at least five years of hands-on experience in data science, combined with proven capabilities in building autonomous agentic workflows within the AWS ecosystem and Enterprise Agent Platforms, we have an exciting opportunity for you! As a Predictive & Agentic AI Engineer, you will bridge the gap between traditional predictive modeling and state-of-the-art generative AI architecture. You will lead the development of advanced data-driven models and design autonomous, multi-agent workflows that solve complex business challenges, automate intricate workflows, and drive deep insights for our MLS subscribers.

Requirements

5+ years of hands-on experience in data science, analytics, machine learning, and AI engineering.
Expert in SQL and advanced SQL, and highly proficient in production-grade Python.
Proficient in frameworks, tools, and libraries such as TensorFlow, PyTorch, scikit-learn, Pandas, and Jupyter Notebook.
Proven experience building stateful, multi-agent workflows using LangChain, LangGraph, CrewAI, LlamaIndex Workflows, or the Microsoft Agent Framework.
Hands-on experience using enterprise-grade agent development and orchestration platforms, specifically Dataiku (utilizing Dataiku LLM Mesh & Agent Hub) or Google Cloud's Gemini Enterprise Agent Platform (formerly Vertex AI Agent Builder) to build, run, and govern production-ready AI agents.
Solid proficiency with building API integrations, advanced Function Calling, and working with the Model Context Protocol (MCP) to seamlessly connect LLMs to external databases and software tools.
Strong data manipulation, wrangling, and mining skills paired with a deep understanding of Vector Databases (e.g., Pinecone, Qdrant, Milvus) and advanced Retrieval-Augmented Generation (RAG and GraphRAG).
Deep understanding of LLM vulnerabilities, including prompt injection mitigation, jailbreak prevention, and the implementation of safety guardrails (e.g., Guardrails AI, NeMo Guardrails).
Expert in AWS artificial intelligence/machine learning services, specifically AWS SageMaker and AWS Bedrock, Bedrock AgentCore, alongside Redshift, Comprehend, and Lex.
Familiarity with LLM tracing and observability infrastructure (e.g., LangSmith, Langfuse, or Arize Phoenix) to monitor and debug complex agent loops.
Strong analytical mindset with excellent communication skills to convey intricate technical concepts and agent behaviors effectively to both engineering teams and non-technical business stakeholders.
Bachelor's degree in Computer Science, Data Science, Mathematics, Statistics, or a related quantitative field.

Nice To Haves

Master’s degree or a Ph.D. in a quantitative field is a plus.
AWS Certified Generative AI Developer – Professional certification or equivalent hands-on experience building autonomous agentic systems using Amazon Bedrock Agents and Bedrock Flows.
Experience with serverless deployment architectures, including AWS Lambda or Glue jobs.
AWS Certified Solutions Architect or AWS Certified Machine Learning Specialty is a plus.

Responsibilities

Work with large and complex real estate data sets to extract meaningful insights and solve a wide array of challenging problems using advanced statistical, machine learning, and large language model (LLM) approaches.
Apply predictive and quantitative analysis, data mining, and experimentation to develop strategies for our product, reporting, and research teams.
Innovate, define, understand, design, and build prototypes of traditional ML models and autonomous, multi-agent AI systems to drive our product roadmap.
Architect and deploy robust AI Agentic workflows capable of task planning, recursive reasoning, tool usage, and self-reflection to automate complex real estate processes.
Hands-on development of end-to-end MLOps and LLMOps pipelines, including dataset curation, model training, prompt engineering, agent trace evaluation, testing, and production deployment.
Continuously refine and optimize traditional ML models and agentic architectures (e.g., via prompt optimization frameworks like DSPy) to maximize accuracy and minimize hallucinations.
Develop methodologies for evaluating both predictive model performance and generative agent compliance, utilizing A/B testing, cross-validation, and LLM-as-a-judge evaluation frameworks.
Partner with Product, Engineering, Research, and other cross-functional teams to translate business needs into scalable, secure AI-driven solutions.
Stay up-to-date with cutting-edge MLOps, LLMOps, and Agentic AI technologies.