- Strong software engineering background with experience building and operating production systems.
- Deep practical experience building applications with LLMs and agentic architectures.
- Strong experience with LangGraph, LangChain, Langfuse, or comparable frameworks.
- Experience designing and implementing LLM evaluations and evaluation pipelines.
- Strong understanding of agentic AI, workflows, tool calling, RAG, structured outputs, context engineering, and graph-based architectures.
- Experience taking AI systems from prototype to production.
- Strong understanding of observability, reliability, latency, and cost considerations in production AI.
- Experience working with multiple LLM providers and understanding their respective trade-offs.
- Ability to move quickly from an ambiguous problem to a working prototype.
- Strong product and business mindset. You care about solving the right problem, not just using the newest technology.
- Ability to communicate technical concepts clearly to clients, executives, product teams, and engineers.
- Experience leading and developing engineers while remaining technically hands-on.
- Strong written and spoken English.

We are looking for someone able to

- Lead, mentor, and develop a team of AI Engineers.
- Remain hands-on, contributing directly to architecture, prototyping, and production development.
- Define technical standards, patterns, processes, and best practices for AI development across Light-it.
- Review architectures and implementations and help the team make strong technical decisions.
- Identify gaps in the team and participate in hiring and developing AI engineering talent.
- Help define the roadmap and priorities of Light-it's AI practice.

Build Production AI Systems

- Architect and implement agentic AI systems, workflows, and LLM-powered applications.
- Design systems using frameworks and technologies such as LangGraph, LangChain, and Langfuse.
- Build reliable agent architectures involving tool calling, structured outputs,



memory, state management, RAG, and multi-step workflows.
- Design robust evaluation systems and eval pipelines to measure quality, reliability, safety, and performance.
- Implement observability and tracing for production AI systems.
- Work on context engineering, prompting strategies, routines, and graph-based agent architectures.
- Make informed decisions about when to use agents, deterministic workflows, traditional software, or combinations of them.
- Optimize AI systems for reliability, latency, and cost.

Work With Clients and Stakeholders

- Participate directly in conversations with clients about AI strategy, opportunities, and implementation.
- Translate complex AI concepts into language that business and product stakeholders can understand.
- Work with stakeholders to identify problems and determine where AI can create meaningful value.
- Turn ambiguous requirements into technical approaches, prototypes, and production systems.
- Help scope AI initiatives and communicate technical trade-offs, risks, timelines, and opportunities.
- Demo AI capabilities and prototypes to clients and internal stakeholders.
- Act as a technical authority for AI during discovery and sales conversations when needed.

Build AI Responsibly in Healthcare

- Design AI systems with healthcare privacy, security, and reliability requirements in mind.
- Understand the implications of working with PHI and regulated healthcare environments.
- Design appropriate safeguards, human-in-the-loop mechanisms, fallbacks, and escalation paths.
- Consider hallucination risk, data privacy, model behavior,



and observability when designing production AI systems.
- Work closely with Engineering and DevSecOps to ensure AI systems meet Light-it's security and compliance standards.
- Continuously evaluate emerging models, frameworks, tools, and approaches.
- Rapidly prototype new AI capabilities through Light-it's Innovation Lab.
- Experiment with new approaches to agents, workflows, context engineering, graph engineering, and human-AI collaboration.
- Evaluate models and providers including OpenAI, Anthropic, Google, AWS Bedrock, and open-source alternatives.
- Explore technologies such as MCP and new standards for connecting agents, tools, and systems.
- Separate technologies that create real value from hype.

Build an AI-Native Engineering Organization

- Help define how software engineering changes in an agentic world.
- Develop and improve Light-it's internal engineering harness and AI development workflows.
- Experiment with coding agents and tools such as Claude Code, Codex, GitHub Copilot, and emerging alternatives.
- Create reusable routines, agents, context, tools, and workflows that improve engineering productivity and quality.
- Help Engineering teams adopt AI effectively rather than limiting AI knowledge to the dedicated AI team.
- Measure the impact of AI-assisted engineering and continuously improve our internal practices.

Other valuable skills

- Experience building AI products in healthcare or another regulated industry.
- Experience with HIPAA and systems handling PHI.
- Experience with AWS and Amazon Bedrock.
- Experience with MCP and agent/tool infrastructure.
- Experience implementing AI-assisted software development workflows at an organizational level.
- Experience with vector databases, knowledge graphs, GraphRAG, or advanced retrieval architectures.
- Experience contributing to technical discovery, solution architecture, or pre-sales conversations.

#J-18808-Ljbffr

📌 Head of AI (Buenos Aires)
🏢 Light-It
📍 Buenos Aires