Note: The job is a remote job and is open to candidates in USA. Tuio is an insurtech company using artificial intelligence to transform insurance operations and build products with real-world impact. The Forward Deployed AI Engineer will work directly with business teams to design, build, evaluate, deploy, and operate agent-based AI systems, taking use cases from prototype to pilots with real users.
Responsibilities
- You will work directly with business teams to understand the processes, decisions, systems, data, exceptions, and risks involved in each use case
- You will design the architecture of agent-based systems, selecting the most appropriate models, patterns, and components for each problem
- You will build applications powered by language models using context engineering, tool calling, structured outputs, RAG, and multi-step workflows
- You will experiment with models, instructions, and architectures to optimize the balance between quality, latency, cost, and reliability
- You will connect agents to internal and external tools through APIs, MCP servers, databases, and SaaS platforms
- You will rapidly create prototypes and evolve them into solutions used with real-world data, systems, and use cases
- You will design evaluations, test cases, and regression tests, and instrument agents to monitor their traces, errors, latency, and costs
- You will implement reliability and control mechanisms such as permissions, validations, retries, fallbacks, action limits, and escalation to a human
- You will deploy and operate solutions using testing, version control, CI/CD, and sound engineering practices
- You will turn the lessons learned from each deployment into reusable components, evaluations, and patterns for future Tuio agents
- You will be involved from the discovery stage and work directly with the people who understand and operate each process. Depending on the project, you will collaborate with AI Managers and colleagues from Business, AI, Data, and Engineering to define and evolve the solution
- You will take end-to-end technical ownership, including AI architecture, code, integrations, evaluations, deployment, observability, and reliability. In some projects, you will also play a significant role in discovery and functional design
- You will build small initial versions to validate hypotheses with real users, measure outcomes, and iterate quickly. We will consciously accept technical debt when it helps us learn, but we will strengthen the architecture, testing, and security before scaling
- You will treat agents as production systems: you will version instructions, tools, and evaluations; analyze traces and failure modes; measure quality, latency, and cost; and test every change against representative cases
- You will adapt the speed of development and level of control to the degree of risk. In low-risk cases, you will experiment with a high level of autonomy. In more sensitive processes, you will introduce permissions, traceability, validations, action limits, data isolation, and human oversight
- You will turn the lessons learned from each project into reusable components, evaluations, and patterns, helping make every new agent faster and more reliable to build than the previous one
Skills
- A university degree in Engineering, Mathematics, Physics, or another related scientific, technical, or quantitative discipline
- At least five years of experience in AI Engineering, Applied AI, Software Engineering, Product Engineering, Data, or a similar technical role
- A strong programming foundation, particularly in Python
- Experience building applications with language models and an understanding of concepts such as context engineering, tool calling, structured outputs, RAG, evaluations, and human-in-the-loop systems
- Experience working with APIs, databases, authentication, testing, Git, and cloud-deployed services
- The ability to turn an ambiguous problem into an architecture and an initial implementation without needing a fully finalized specification
- The ability to use data, metrics, traces, and evaluations to analyze and improve the behavior of an AI system
- Strong engineering judgment when balancing quality, latency, cost, security, maintainability, and development speed
- An interest in working closely with users and the ability to communicate clearly in English with both technical and business stakeholders
- Experience deploying agents or language-model-based products used by real users
- Experience with tool calling, RAG, MCP servers, vector databases, workflows, or agent architectures
- Experience building evaluation, observability, monitoring, or human-in-the-loop systems for AI applications
- Knowledge of cloud infrastructure, containers, CI/CD, queues, or event-driven architectures
- Experience using AI-assisted development tools to build, test, and maintain software
- Experience working with sensitive data, access controls, traceability, or AI systems in regulated environments such as insurance or insurtech
Benefits
- Flexible working hours
- Remote work
- A growth plan within a rapidly expanding company
- The opportunity to work at a company where AI is not a trend, but a fundamental part of how we build products and improve our work every day
- The opportunity to build solutions with real impact
- Flexible compensation
- Health insurance
- Tuio insurance policies
- Home-office equipment
- Team-building events
Company Overview