Senior AI Software Engineer

Own the AI agents that make engineers, quants, and traders faster

You’ll design and ship real, production AI agents that cut through operational overhead and give developers, quants, and traders their time back.

This is not a “labs” role. You’re building autonomous and semi-autonomous systems that touch codebases, APIs, internal tools, and data pipelines-safely, securely, and at scale.

You own architecture. You own implementation. You own the behavior of agents that must actually work in the wild.

If you’ve been itching to push beyond toy demos and build serious agentic systems in production, we want to talk to you.

What You’ll Work On

You’ll sit at the intersection of AI systems engineering and workflow automation, turning messy real-world processes into dependable AI-powered workflows:

• Autonomous AI agents executing multi-step workflows across internal systems
• Retrieval-augmented generation (RAG) architectures using structured, permissioned data
• Agent memory and orchestration layers
• Inter-agent communication patterns
• Tool schemas that drive deterministic agent behavior
• Safeguards for LLM failure modes and constraints
• Automation that supports developers, quants, and traders
• Evaluation and integration of new AI models and frameworks

You’ll be working with Python at an expert level, RAG patterns and vector databases (e.g., Pinecone, Chroma, pgvector), and agent frameworks such as LangChain or LangGraph.

Your job: turn ambiguous workflows into robust agent systems that run with minimal human babysitting.

What You’ll Be Doing

• Designing and implementing autonomous AI agents that execute multi-step workflows across internal systems
• Building and productionizing RAG architectures on top of structured, permissioned data sources
• Developing durable agent memory and orchestration layers for complex workflows
• Creating inter-agent communication patterns so multiple agents can coordinate effectively
• Defining tool schemas and translating ambiguous stakeholder needs into deterministic agent behaviors
• Implementing safeguards against hallucination, prompt drift, context issues, and rate limits
• Partnering directly with developers, quants, and traders to uncover high-value automation opportunities
• Deploying AI agents into production environments and iterating based on real-world feedback
• Evaluating and integrating emerging models, frameworks, and open-source AI tooling into existing systems
• Designing systems to operate reliably with minimal human intervention
• Ensuring sensitive and proprietary data is handled and integrated securely within AI workflows, senior software engineer, software engineer, AI engineer, AI systems engineer, machine learning engineer, agentic systems, AI agents, autonomous agents, workflow automation, Python, async Python, LLM, large language models, RAG, retrieval-augmented generation, vector database, Pinecone, Chroma, pgvector, LangChain, LangGraph, orchestration layer, inter-agent communication, prompt engineering, hallucination mitigation, prompt drift, context management, rate limiting, secure data handling, production AI, trading technology, quant tools, developer productivity, internal tools

• 5-10+ years of software engineering experience
• Proven track record building AI-powered applications, automation platforms, or agentic systems in production
• Demonstrated ability to design systems that operate with minimal human intervention
• Expert-level Python, including async, typing, packaging, and testing best practices
• Experience building AI agents that interact with APIs, tools, codebases, and shell environments
• Strong understanding of RAG patterns and vector databases such as Pinecone, Chroma, or pgvector
• Familiarity with agent frameworks like LangChain, LangGraph, or similar tools
• Deep understanding of LLM behavior, constraints, and mitigation strategies
• Strong system design skills and comfort working independently in a fast-evolving space
• Experience handling sensitive or proprietary data securely

The Experience That Will Really Get Our Attention

You’ve moved beyond chatbot front-ends and have actually wired LLMs into serious back-end systems-agents that call tools, inspect code, touch production-like data, and keep running without constant human guardrails.

You’re comfortable owning an ambiguous problem end to end: mapping workflows, designing agent behaviors, choosing the right RAG and vector patterns, implementing safeguards, and hardening everything for production.

High-signal experience: Python (async) * RAG * vector databases (Pinecone * Chroma * pgvector) * LangChain * LangGraph * AI agents * LLM safety * orchestration layers * secure data handling