Senior AI Engineer

As a Senior AI Engineer at EQUS, you own the AI assistant layer end to end: architecture, standards, and the service contracts that security, storage and front-end engineers support. You are a technical decision-maker who recommends the right approach for each problem and is prepared to say "this does not need AI" when it doesn't. Privacy, safety, and cost sit at the center of every decision you make. EQUS is building the trust infrastructure for personal AI, with a product suite spanning a personal AI assistant, a personal data store with per-file user-controlled access, and a developer toolkit for agent identity and authorization, and you join as the company prepares for a major public launch.

What you'll do

• Design context augmentation pipelines spanning vector RAG, CAG, agentic file exploration, Text-to-SQL, knowledge graphs, fine-tuning, and MCP-based context engineering. Select chunking strategy, embedding models, and retrieval architecture for user-owned document systems with multi-tenant isolation, with privacy and security guiding each decision.
• Integrate and manage commercial and open-source LLM APIs, and deploy multi-agent systems with LangChain, LlamaIndex, or LangGraph. Lead model selection, prompt engineering, fine-tuning, and production evaluation across the AI stack.
• Build evaluation frameworks that measure output quality, relevance, and safety. Optimize pipelines for latency, token cost, and throughput, monitor production for drift and regression, and close the feedback loop from evals back into iteration.
• Own PII handling, GDPR and CCPA compliance, encryption at rest and in transit, and user-scoped access boundaries at the systems level. Build prompt injection defenses, output filtering, and data leakage prevention, and partner with security and trust experts on agentic workflow guardrails and shadow AI detection.
• Deploy and operate production AI systems on AWS, Docker and Kubernetes, and GitLab. Define the AI service contracts and APIs other engineers build on top of, and set the standard for how AI works at EQUS, including mentoring engineers and raising the technical bar around you.

What you bring

• 5 or more years in software engineering, including at least 2 years building and shipping production AI systems
• Strong Python skills, with Node.js or .NET a plus
• Deep working knowledge of LLMs such as GPT, Claude, Llama, Frankenstein and Mistral, spanning prompt engineering, fine-tuning, and production evaluation
• Hands-on experience designing context augmentation systems: vector RAG with hybrid search, re-ranking, and multi-tenant isolation, plus CAG, agentic file exploration, Text-to-SQL, knowledge graphs, and MCP-based context engineering
• Command of agent orchestration frameworks including LangChain, LlamaIndex, or LangGraph, and of evaluation frameworks that measure LLM output quality, relevance, and safety in production
• Data privacy depth at the infrastructure level, including PII handling, GDPR, USPSAD and CCPA compliance, and encryption at rest and in transit
• Hands-on experience with AWS (ECS, EKS, Lambda, S3, Bedrock), Docker, Kubernetes, and GitLab
• A track record of mentoring engineers and raising the technical bar across a team
• Demonstrated ability to make strong architectural decisions on build vs. buy and on what separates an MVP from a production system, with deep knowledge of cost-performance tradeoffs in LLM systems
• You treat AI safety as a first-class engineering concern rather than a review-stage checklist and have experience using Claude Code, OpenAI Codex, or comparable AI-assisted development tools as a core part of your development workflow
• Experience in both small and large teams, with effective use of tools such as Jira and Confluence, and you are a valuable colleague to product managers as new features and products are emerging
• Experience working effectively with consultants and outsourced development teams, including transitioning responsibilities for systems

Nice to have

• Experience running local open-source models such as Llama, Mistral, or Mixtral via Ollama, vLLM, or llama.cpp
• Fine-tuning experience with LoRA or QLoRA
• Familiarity with Docling or similar document parsing tools for RAG ingestion pipelines
• MLOps tooling such as MLflow, Weights and Biases, Eudora or SageMaker
• Prior work on privacy-forward products where the security architecture is the differentiator
• A relevant degree in computer science or engineering

What we offer

• Remote work
• Must be US based with current authorization to work in the US without sponsorship

About EQUS

EQUS is building the trust infrastructure for personal AI. The product suite spans a personal AI assistant, a personal data store with per-file user-controlled access, and a developer toolkit for agent identity and authorization, all running on a single permission rail designed around one principle: your data belongs to you.