The posting
Senior AI Engineer Remote · Full-Time - Colombia
About Cadre AI
Cadre AI is an AI strategy and integration firm that builds production AI systems for B2B companies in private equity, wholesale lending, real estate, and SaaS. We do not build decks about what AI could do. We ship systems that move revenue, compress costs, and automate the work that used to take entire teams.
The Role
You are the technical backbone of Cadre AI's delivery engine. While AI strategists own the client relationship and translate business problems into requirements, you own the build: complex backend AI modules, agentic workflows, retrieval systems over messy unstructured data, and the infrastructure that keeps them running at scale. There are no over-the-wall handoffs here. You co-develop requirements in real time alongside the strategist, which means what gets built is precise, purposeful, and shipped fast.
You thrive on deep, uninterrupted technical execution. You are not a generalist who dabbles in AI. You have built production LLM systems, tuned RAG pipelines until retrieval accuracy became a non-issue, and debugged agentic chains under pressure. You know the difference between a prototype and a production-grade system, and you build the latter by default.
Beyond client delivery, you look for patterns. When three clients ask for similar components, you abstract the best version into an internal tool that speeds up every future delivery. That force-multiplier instinct is what separates this role from a traditional contractor engagement.
What You'll Do
Production AI System Development Design and build complex backend AI modules: custom agentic workflows, multi-step reasoning chains, and orchestrated tool-use pipelines across client environments. Translate architectural blueprints from AI strategists into high-performance, production-ready code with clear error handling, observability hooks, and latency targets. Own end-to-end delivery of AI components from initial scaffold through deployment, ensuring the system behaves exactly as designed in the client's actual environment. Use AI coding assistants as a standard part of your workflow to accelerate output without sacrificing code quality or architectural integrity. Advanced RAG and Retrieval Infrastructure Architect and optimize RAG pipelines over unstructured, messy, and multi-modal data sources, applying chunking strategies, embedding models, and re-ranking techniques that move retrieval accuracy from acceptable to precise. Implement vector database solutions based on query patterns, data volume, and latency requirements specific to each engagement. Diagnose and fix retrieval failures in production: identify whether the problem is in chunking, embedding, indexing, or prompt construction, and fix the right thing. Agentic Orchestration and Context Engineering Build and maintain agentic systems with particular attention to tool selection, memory architecture, and failure recovery. Manage context window constraints deliberately: design systems that stay within limits, degrade gracefully when they approach them, and never silently drop critical information. Optimize prompt pipelines for latency and cost without degrading output quality, benchmarking before and after each change with data rather than intuition. Infrastructure, Deployment, and Quality Own deployment of AI services using modern IaC tooling and containerized environments, ensuring repeatable, auditable deploys across dev and production. Write tests at the right level of the pyramid: unit tests for deterministic logic, integration tests for pipeline behavior, and evaluation harnesses for non-deterministic AI outputs. Maintain production systems with structured logging, alerting, and cost monitoring so issues surface before clients notice them. Internal Tooling and Platform Abstraction Identify patterns across client engagements and abstract the best implementations into reusable internal tools, libraries, and services that reduce future build time. Document internal tools clearly enough that another engineer can use them without asking you for help, treating internal documentation as a first-class deliverable. Actively contribute to Cadre's internal AI platform, participating in architecture reviews and proposing improvements when you see a better way.
What You Need to Succeed 3+ years of backend engineering experience, with at least 1 years building and shipping production AI agents in a client-facing or fast-paced product environment, not just internal tooling or research prototypes. Deep, hands-on experience with vector databases, advanced RAG optimization, and agentic orchestration. You have specific examples of where you improved retrieval accuracy or reduced pipeline latency in a live system. Proficiency in Python and Node.js with strong command of backend architecture patterns: API design, async processing, parallel execution, and secure credential handling. Infrastructure fluency: you use IaC tools (Terraform or Pulumi), deploy with Docker and Kubernetes, and understand the difference between a deployment that works and one that is maintainable at scale. Systems design instincts that favor abstraction and reuse. When you build something complex for one client, your first thought is how to make it generic enough to deploy again. Certified Claude Architect or the demonstrated ability to earn the certification within your first 60 days. You are a Claude expert or you are actively becoming one. AI-native working style: you use coding assistants as a force multiplier, not as a crutch, and you know when to trust generated code and when to rewrite it from scratch. Complexity tolerance that others would describe as unusual. You do not avoid unstructured data problems or legacy integration constraints. You find them interesting. Ownership orientation: you finish what you start, you flag blockers early rather than late, and you take pride in the quality of systems that have your fingerprints on them.
Why Cadre AI
Real ownership. You run engagements end-to-end. No handoffs mid-project, no waiting on approvals. You own the outcome. Variety and velocity. Every engagement is a different industry, a different problem, and a new chance to build something that matters. You won’t get bored. AI-native culture. We don’t just build AI for clients. We use it to run our own operations. You’ll work with people who are as obsessed with the tools as you are. Access to the frontier. Through partnerships with Anthropic and OpenAI, you’ll be among the first to experiment with new models and capabilities. Upside. We’re bootstrapped, profitable, and growing fast. Early team members share in the success they help create. No bureaucracy. Small pods. Clear accountability. The best idea wins, regardless of who says it.
If you have built production AI systems that other engineers point to as examples of how it should be done, and you want to do that work across a portfolio of complex, high-stakes client problems, this is the role.
