• Design, build, and maintain real-time AI-powered APIs and services in Python. • Architect, optimize, and deploy LLM workflows leveraging LangGraph or similar orchestration frameworks. • Develop agentic logic and reasoning pipelines for conversational and task-based automation. • Build scalable infrastructure to support multi-model, low-latency, and streaming AI experiences. • Collaborate with backend and product teams to translate complex customer workflows into reliable AI systems. • Ensure high performance, reliability, and data security across production workloads. • Deploy and manage applications on modern cloud environments (AWS / GCP). • Implement monitoring, logging, and fault-tolerance mechanisms for production-grade AI systems. • Contribute to architectural and product decisions, balancing innovation, safety, and scalability.

• Design, build, and ship production-grade AI agents for meditation, wellbeing, mental health, sleep, anxiety, stress, and behavior change. • Build agentic workflows that combine technical reliability with clinical and evidence-informed reasoning. • Integrate AI agents with backend APIs, data systems, content systems, and product surfaces. • Build evaluation, testing, and monitoring systems for reliability, safety, quality, and continual improvement. • Design clinical and wellbeing evaluation rubrics for AI outputs, including clarity, usefulness, safety, evidence alignment, and risk. • Work with the Head of AI, clinical research team, product, and engineers to translate clinical and research requirements into product and system design. • Identify risks in AI-generated wellbeing outputs, including unsupported claims, unsafe recommendations, overstatement, inappropriate advice, and poor handling of vulnerable users. • Improve prompts, tools, agent workflows, and retrieval systems based on both engineering performance and clinical/product quality. • Architect flexible, modular workflows that make it easy to test, extend, and improve AI capabilities. • Help define when agentic systems should answer, ask for more context, escalate, refuse, or redirect. • Design data and feedback pipelines that improve agent performance through real-world usage, evaluation, and, where appropriate, fine-tuning. • Translate early AI prototypes into scalable, maintainable product systems used by millions of people. • Help shape Insight Timer’s internal standards for building safe, evidence-informed AI wellbeing products.

Role Description Buscamos un AI Engineer part-time desde LATAM para unirse a uno de nuestros proyectos internos de mayor crecimiento — un sistema de automatización basado en agentes de IA con integración en plataformas de pago. - Desarrollar agentes de IA autónomos con capacidad de toma de decisiones y ejecución de tareas financieras. - Construir e integrar APIs REST para conectar el agente con sistemas externos. - Implementar flujos de trabajo agenticos usando LLMs (OpenAI, Anthropic). - Trabajar con protocolos emergentes de comunicación entre agentes (MCP, A2A). - Colaborar directamente con el equipo técnico en decisiones de arquitectura. Qualifications - Python sólido y experiencia con APIs REST. - Experiencia práctica con LLMs y frameworks de agentes (LangChain, LlamaIndex, OpenAI API). - Conocimiento de arquitecturas agenticas y flujos autónomos. - Trabajo asíncrono y autónomo — sabemos gestionar sin micromanagement. - Español nativo o fluido. - 20 horas semanales de disponibilidad real. Requirements - Imprescindibles: - Python. - LLMs / Large Language Models. - LangChain. - OpenAI API. - REST API Development. - AI Agents / Agentic AI. - Prompt Engineering. - Deseables: - Anthropic / Claude API. - LlamaIndex. - MCP (Model Context Protocol). - Stripe API. - FastAPI. - Fintech. Benefits - Proyecto con impacto real desde el primer día. - Equipo pequeño, decisiones rápidas. - 100% remoto. - Compatible con horario LATAM. - Posibilidad de crecer con el proyecto a largo plazo.

• Design, build, and maintain agentic AI workflows, owning the full lifecycle from prototype through production monitoring. • Build and operate generative media pipelines using ComfyUI, Fal, and direct model APIs. • Ship internal AI tools and dashboards that give Post, Art, Content, and Marketing teammates direct access to AI workflows without engineering support. • Build evaluation, monitoring, and failure-recovery systems for every workflow you deploy. • Partner with the CTO and Tech team on infrastructure, deployment, security, and IP and data handling standards across all AI systems. • Translate department pain points into engineered solutions with measurable outcomes.