At Unboxed we help companies develop skills at the pace of change. We leverage our advanced technology, engaging content, and years of experience with industry leaders to empower employees to upskill, improve, and thrive in a rapidly changing business world.

We are looking for a AI Software Engineer to join our software development team and focus on building and scaling our next-generation SaaS product. This role centers on designing and implementing advanced, production-grade multi-agent AI systems that power intelligent features and workflows for our education and training platform and will work closely with product and engineering teams to deliver scalable, reliable, and extensible AI-driven solutions that enhance user learning experiences.

Key Responsibilities:

• Design, develop, and maintain multi-agent AI pipelines using LangChain and LangGraph to orchestrate complex, stateful workflows supporting educational content delivery, assessment, and personalization.
• Architect and implement structured data flows with Pydantic for robust state management and data validation between agents and workflow stages.
• Build graph-based execution models to enable parallel and sequential processing of tasks, optimizing for scalability and responsiveness in a SaaS environment.
• Integrate Azure OpenAI services and other Azure cloud components to deliver advanced AI features within the product.
• Ensure system reliability through robust error handling, retry logic, and clear state transitions.
• Extend and adapt AI systems to incorporate new agent types, validation steps, or SaaS features as the platform evolves.
• Collaborate with cross-functional teams—including product, UX, and DevOps—to deliver high-quality, production-ready features.
• Contribute to documentation, telemetry, and monitoring to support ongoing product improvement and user analytics.

Requirements

Requires 1-3 years of professional experience with:

• LangChain (v0.3.6+) for LLM application development
• LangGraph (v0.2.43+) for multi-step AI workflow construction
• Pydantic (v2.10+) for data validation and settings management
• Agentic system design and multi-agent pipeline orchestration
• Experience integrating with Azure OpenAI and deploying cloud-based AI services.
• Strong programming skills and experience building scalable, stateful AI systems.

These skills are preferred, but not strictly required:

• Experience or education in data science or machine learning
• Experience with Azure Functions, Azure Search (vector and hybrid search), and Azure Cognitive Services.
• Familiarity with data manipulation libraries such as pandas and numpy.
• Understanding of vector embeddings, similarity search, and Retrieval-Augmented Generation (RAG) patterns.
• Experience working with educational or training data validation.
• Proficiency with tools like MkDocs (documentation), Plotly (data visualization), and NetworkX (graph analysis).
• Experience with DevOps/MLOps practices, including Azure DevOps pipelines, telemetry (OPIK tracing), model deployment, and versioning.

NOTE: We prefer candidates who are within a 3-hour drive (150-200 miles) of Richmond, VA.

Benefits

Unboxed team members benefit from our comprehensive compensation and rewards program which includes:

• Competitive salary and benefits
• Ample paid time off – 5 weeks PTO (pro-rated in 1st year), 6 paid holidays plus winter break between 12/26 and 1/1
• Dynamic and convenient office location – unlimited snacks, casual dress code, covered parking and gym on site
• Open communication and a commitment to fostering teamwork across the organization