AI Developer (Contract)

AI Developer

Eleven Systems LLC | Eleven Family Office (Eleven)

Location: Remote

Employment Type: Contract

Company Description

Eleven Systems develops Eleven Family Office (EFO), an advanced software platform designed to streamline the complex needs of Ultra High Net Worth (UHNW) families. The platform provides unified tools for property oversight, investment tracking, secure messaging, family logistics, and concierge service management. Built for scale, security, and simplicity, EFO supports family offices, concierge teams, and wealth management firms.

For more information, visit www.elevensystems.com.

About the Role

Eleven Systems is hiring an experienced AI Developer to build and enhance the agentic AI assistant within our property and financial management platform.

Eleven AI is a Java-based service that transforms user requests into answers and actions through a supervisor-worker architecture of LLM agents. Supervisors route requests to specialized workers, workers execute tools against our backend services, and results stream back to the client.

The platform also includes document-understanding agents for travel itineraries, invoices, identity documents, and insurance policies; cross-domain request routing; a sandboxed Python runtime for reports and spreadsheets; encrypted chat memory; multi-tenant security; usage and cost tracking; and a Langfuse evaluation suite.

You will own features end to end, from technical design through implementation, testing, and merged pull requests, using Claude Code as a daily engineering partner.

What You Will Do

• Design, develop, and deploy AI agents, supervisors, workers, tools, and prompts using LangChain4j and Spring Boot.
• Integrate and optimize OpenAI, Anthropic, and AWS Bedrock models, including structured output, prompt caching, fallbacks, timeouts, and per-tenant configuration.
• Build document-processing pipelines for PDF and image ingestion, vision-based extraction, verification, classification, and normalization.
• Develop AI workflows that securely retrieve information and execute authorized actions against backend services.
• Improve AI response quality using evaluation datasets, Langfuse traces, automated testing, and application-level output validation.
• Implement safeguards against prompt injection, unauthorized tool execution, cross-tenant data leakage, and fabricated links or identifiers.
• Develop and maintain secure, scalable Spring Boot APIs and streaming responses.
• Prepare concise technical design documents, write automated tests, and deliver clear, reviewable pull requests.
• Collaborate with engineering and QA teams to troubleshoot issues, improve performance, and deliver production-ready functionality.

Our Technology Stack

Java 21, Spring Boot 3.5, LangChain4j 1.20 (agentic module and MCP client), OpenAI, Anthropic, AWS Bedrock Converse, MongoDB 7 with field-level encryption, SSE and WebSocket streaming, Apache Tika, PDFBox, OpenTelemetry, Langfuse, Docker, Jenkins, GitLab, Postman/Newman, and Claude Code with project-defined agents and skills.

What We Are Looking For

• 3+ years of professional backend Java development experience.
• 2+ years of production experience with Spring Boot, including REST APIs, asynchronous execution, thread pools, and request context propagation.
• 1+ years of hands-on experience delivering LLM-powered applications, including tool calling, multi-agent orchestration, structured output, prompt design, memory, and context management.
• Strong understanding of LLM limitations and failure modes, with experience implementing code-level safeguards and automated evaluations.
• Production experience with MongoDB, including document modeling, indexes, TTL, and aggregation.
• Proficiency using Claude Code or similar AI coding agents in real software development projects, including reviewing and correcting AI-generated code.
• Strong security awareness, including secrets management, encryption at rest, tenant isolation, and AI-specific security threats.
• Ability to independently design, implement, test, and deliver production-ready software.
• Clear written communication through technical design documents, pull requests, and project documentation.

Nice to Have

• Hands-on experience with the LangChain4j agentic module or Model Context Protocol (MCP).
• Experience with AWS Bedrock Converse, including inference profiles, structured output, and prompt caching.
• Familiarity with LLM evaluation platforms such as Langfuse, LangSmith, or Braintrust, including LLM-as-judge evaluation methods.
• Experience with document AI, streaming APIs, and vision-based extraction.
• Python data processing experience with pandas, openpyxl, or matplotlib.
• Experience with secure code execution or sandboxing technologies such as gVisor.
• Knowledge of property management, family office software, financial services, or fintech applications.

How We Work

Claude Code is an integral part of our engineering workflow. We use a structured development process that includes planning, technical design, implementation, parallel code review, security review, automated test generation, QA traceability, and delivery through Conventional Commits and GitLab merge requests.

Our repositories include project-defined AI agents and skills that establish engineering conventions, with CLAUDE.md serving as the project's development guide.

Engineers are responsible for approving implementation plans and technical designs, reviewing AI-generated code and findings, and determining what is ready for production.

Expect to spend a significant portion of your day directing, reviewing, and improving AI-generated work. We are looking for developers who understand what high-quality software looks like and know how to use AI development tools effectively to deliver it.