Senior Full-Stack Engineer

Worldwide | Sept. 14, 2026

Report as Closed

Company: LDD

Country: Worldwide

Salary: $42,000 - $48,000

Type: Remote

Employment: Full-time

Description: DataLaw. We are a Chilean compliance tech startup. None

Agent AI

• Experience building agents or functionalities with LLMs in production: tool use, multi-step orchestration, context management and evaluation of results.
• Advanced use of code agents in daily work, including knowing how to configure context, skills and rules for a specific project.
• Criteria for using AI without delegating technical responsibility: you know when to trust the output, when to review it and how test it.
• Knowledge of the agents' own risks (prompt injection, data leakage via tools, unauthorized actions) and how to mitigate them.

Full stack

• Senior experience developing full stack applications in production, with end-to-end responsibility for the functionalities you build.
• Proficiency in TypeScript; experience with Angular (or willingness to work with it) and Python/FastAPI, with judgment in state and performance management.
• Experience deploying and operating services in GCP (Cloud Run, Cloud SQL), with Docker and Terraform in daily work.
• Practical knowledge of B2B multi-tenant SaaS: data separation, authorization and traceability.
• Design of resilient and secure systems: failure prevention, recovery from errors and layers of defense against external threats.

At LeyDeDatos we build the software with which companies manage their data protection obligations: registration of processing activities, consents, rights of the owners and classification of data with AI. The new Chilean regulation is forcing thousands of organizations to organize their processes now, and we build the tool with which they will do it. What goes into production today is used by real customers tomorrow.

The technical challenge is a good one: a multi-tenant B2B SaaS where AI agents operate on personal information from different organizations. Isolation between tenants, permissions and traceability are part of the design of everything, including agents. Here agentic AI is not an experiment: it is a product.

We work with product and engineering very closely, and each person has real ownership over what they build and a voice in the decisions. Whoever comes in now will help define how we build software and, over time, set its technical direction.

There are two sides to the role, and they both matter:

• IAgent in the product. Our product classifies personal data, automates compliance flows and assists users with agents. You are going to design those agents: what tools they expose, how they are orchestrated, how they are evaluated and what limits they have. Since they touch personal data from several organizations, an agent has to respect permissions, isolation between tenants and traceability just like any other part of the system.
• Agent AI in development. We work with code agents integrated into the daily flow. You will maintain the context that makes these agents work well in our project: Markdown documentation, skills, rules per repository and MCP servers. And you will be responsible for what reaches production, whether it was written by a human or an agent.

On a day-to-day basis you will:

• Build complete functionalities with TypeScript, Node.js, React or Next.js, APIs and persistence, from the definition of the problem to its operation in production.
• Design and integrate agents into the product: tool use, multi-step orchestration, RAG, guardrails, evals and inference cost control.
• Define how agents access personal data respecting permissions, multi-tenant isolation and auditing.
• Maintain the context of the team's code agents (Markdown, skills, rules, MCP).
• Deploy and operate your changes in GCP, considering availability, performance and costs.
• Write tests, review code and contribute to the observability of services and agents.

You will have ownership over relevant functionalities, direct collaboration with the team and room to grow towards broader technical responsibilities.

• Hybrid modality in Santiago, Chile, with flexible hours.
• Initial fixed-term contract and then indefinite. The first period functions as mutual adjustment; The intention is to continue indefinitely.
• Full time.
• Computer provided by the company.

• Have built or exposed MCP servers, or integrated agents with external systems via tools.
• Experience with LLM evaluation frameworks and with agent observability in production.
• Have worked on regulated products or with audit requirements.
• Experience designing authorization

  Apply here:

  Web: Apply here

  Emails: