Agentic AI Systems Engineer

• Senior-level software engineering background, with strong production experience building backend systems and APIs in Python.
• Proven experience deploying AI agents or tool-using LLM systems into production, not only prototypes or demos.
• Hands-on experience with multi-agent or tool-orchestrated architectures, including runtime control, context/state handoffs, and routing decisions.
• Strong experience with RAG systems in production: chunking, indexing, embeddings, retrieval optimization, grounding, and evaluation of retrieval quality.
• Experience implementing evaluation and quality-control workflows for AI systems, such as benchmark suites, LLM-as-a-judge, regression testing, adversarial testing, or production quality monitoring.
• Strong understanding of observability for AI systems, including traces, logs, prompt/model versioning, token usage, runtime diagnostics, and failure analysis.
• Experience designing safe failure modes for agentic systems, such as deterministic fallbacks, constrained responses, bounded retries, or human-in-the-loop escalation.
• Solid API and systems integration experience across enterprise environments.
• Strong English communication skills, with ability to discuss architecture, trade-offs, and production behavior clearly with a distributed international team.
• Preferred / Strong Plus
• Experience working in regulated, high-trust, or compliance-heavy domains such as healthcare, legal, insurance, finance, industrial, or regulatory intelligence.
• Experience working close to the model/provider layer and making pragmatic decisions around model selection, hosting, abstraction level, and production constraints, including open-source or self-hosted model environments.
• Experience with event-driven architectures, workflow systems, or message-based coordination patterns for AI runtime execution.
• Familiarity with AI coding tools and AI-assisted software development workflows that improve speed without sacrificing quality.

• Design and implement production agentic systems that use tools, retrieval, and structured decision-making to automate or augment complex regulatory and expert workflows.
• Modernize older, human-driven, or rule-based business processes into structured agentic architectures, choosing pragmatic patterns such as orchestration, choreography, or hybrid approaches depending on runtime and business constraints.
• Build and optimize RAG systems over large, heterogeneous knowledge sources, including public regulatory data, internal expert content, and customer-provided data, with strong focus on retrieval quality, grounding, and evidence coverage.
• Design robust tool-calling and routing logic, including context handoffs, agent coordination, and fallback behavior when tools, APIs, or retrieval dependencies fail.
• Establish strong observability and quality frameworks for AI systems, including tracing, prompt and model version tracking, token/cost monitoring, failure analysis, regression checks, evaluation pipelines, and drift detection.
• Implement guardrails and safe degradation patterns so systems remain usable and trustworthy under failure, ambiguity, or incomplete context.
• Integrate AI services with internal and external enterprise systems through APIs, event-driven components, workflow services, and secure data access patterns.
• Contribute to an AI-native engineering culture, using AI tools to accelerate delivery while maintaining strong architecture, code review, testing, and production standards.

You will design, build, and operate production-grade agentic AI systems for a global regulatory intelligence company that is transforming from expert-led content and workflow services into an AI-first platform business. Your work will focus on modernizing legacy and human-heavy enterprise workflows into reliable AI systems that combine expert knowledge, regulatory content, and customer-specific data to support internal operations and customer-facing products.

This is not a chatbot-only role and not a prompt-engineering role. Success in this position requires strong software engineering fundamentals, hands-on experience with agents in production, and the ability to design systems that are observable, auditable, controllable, and safe to operate in high-trust environments. The strongest candidates will be comfortable reasoning about agent runtime behavior, tool use, orchestration, retrieval quality, evaluation frameworks, and safe degradation patterns — not just model outputs.

Join Niuro’s 100% remote work model with global flexibility. We invest in your professional development through ongoing training programs and leadership opportunities, ensuring continuous growth. You’ll be part of a global community dedicated to technological excellence, supported by a robust administrative infrastructure that lets you focus on impactful work. Upon successful completion of the initial contract, there may be opportunities for long-term collaboration and full-time employment. You’ll enjoy a collaborative culture that values technical excellence, continuous innovation, and work-life balance, along with opportunities to participate in technically rigorous projects that drive real-world impact.