Senior Software Engineer - AI App Enablement & Observability

EuroPython; Dublin; Hybrid; EUR 120,000 - 180,000; Full time

Job summary

Bloomberg Industry Group is seeking a Platform Engineer to shape scalable AI platforms, tooling, and paved roads that enable engineers to ship reliable, secure, and high-performing systems at scale.

You will collaborate across AI product, infrastructure, and platform teams to build observability, telemetry pipelines, and self-improving infrastructure, driving cost efficiency and end-to-end ownership of critical platforms.

Qualifications

• Experience building production software or platform systems.
• Strong fundamentals in distributed systems or backend platforms.
• Interest or experience in observability and debugging complex systems.
• Interest in AI/ML systems, LLMs, or agent-based architectures.

Responsibilities

• Build and evolve AI platform tooling and developer workflows.
• Design developer-friendly APIs, SDKs, and interfaces.
• Operate observability platforms and telemetry pipelines.
• Implement distributed tracing and instrumentation standards.
• Own end-to-end platform projects from RFCs to production support.
• Collaborate with AI teams to drive adoption of tooling and standards.

Skills

Distributed systems

Backend platforms

Observability

AI/ML systems

OpenTelemetry

Kubernetes

CI/CD tooling

Security awareness

Tools

Prometheus

Grafana

Terraform

Cloud tooling (AWS/Azure)

Job description

Platform Engineering builds the core platforms, tooling, and paved roads that Bloomberg engineers rely on to ship reliable, secure, and high-performing systems at scale.

The AI App Enablement & Observability team accelerates how AI products are built across Bloomberg Industry Group. Our mission is to make AI systems reliable, performant, cost-efficient, and continuously improving through platform tooling, deep observability, and automated feedback loops.

We build developer-facing platforms and workflows that enable teams to experiment, deploy, and operate AI and agent-based systems with confidence. This includes LLM gateways, agent platforms, benchmarking systems, telemetry pipelines, and self-improving infrastructure that closes the loop between observability and action. We emphasise strong developer experience, intuitive APIs/SDKs, and end-to-end ownership.

What’s in it for you?

You will help define how Bloomberg Industry Group builds and operates AI systems at scale by working on platforms that:

• Accelerate AI product development through reusable tooling and paved roads
• Provide end-to-end observability across AI systems (models, agents, pipelines, applications)
• Enable self-improving systems through telemetry-driven feedback loops
• Optimise cost, performance, and reliability of AI workloads
• Support both production AI systems and internal engineering agentsYou’ll collaborate across AI product, infrastructure, and platform teams to deliver foundational systems.

Responsibilities

• Platform & Enablement
  • Build and evolve AI platform tooling (e.g., developer workflows, benchmarking systems)
  • Design developer-friendly APIs, SDKs, and interfaces
  • Contribute to systems across the Model Development Lifecycle (experimentation, deployment, evaluation)
• Observability & Telemetry
  • Build and operate observability platforms and telemetry pipelines (logs, metrics, traces, events)
  • Provide visibility into latency, token usage, cost, quality, drift, and reliability
  • Define instrumentation standards, schemas, and conventions
  • Implement distributed tracing using modern approaches (e.g., OpenTelemetry)
• AI System Insights & Debugging
  • Enable end-to-end debugging of AI and agent workflows (model calls, tool usage, retrieval, orchestration)
  • Build benchmarking, regression detection, and performance analysis capabilities
  • Support observability for both production systems and internal engineering agents
• Closed-loop Optimization & Automation
  • Develop systems that turn telemetry into action (automated experimentation, regression detection, alerting)
  • Build feedback loops that continuously improve model quality and system behavior
  • Enable self-healing and self-optimising workflows
• Cost, Performance & Reliability
  • Build tooling for cost visibility, forecasting, and optimization
  • Define SLOs, alerting, and performance tuning practices
  • Improve reliability and scalability of AI infrastructure
• Own projects end-to-end (RFCs, architecture, implementation, rollout, production support)
• Partner with AI teams to drive adoption of platform tooling and standards
• Produce high-quality documentation and improve developer experience

Requirements

• Demonstrated experience building production software or platform systems
• Strong engineering fundamentals with distributed systems or backend platforms
• Experience or strong interest in observability and debugging complex systems
• Experience or strong interest in AI/ML systems, LLMs, or agent-based architectures
• Strong ownership mindset and ability to drive ambiguous problems to production
• Hands-on experience with modern agentic coding tools (e.g., Claude Code, Codex CLI, Cursor) and multi-model workflows
• Working knowledge of agent architecture internals (context engineering, tool loops, sub-agent orchestration)

Nice-to-Haves

• Experience with OpenTelemetry and modern observability ecosystems, including instrumentation, collectors, exporters, and tools like Prometheus, Grafana, and tracing/log systems
• Experience designing and operating telemetry pipelines, including sampling, retention, cardinality, and cost tradeoffs, as well as integrating observability into CI/CD and developer workflows
• Familiarity with AI/agent frameworks, including instrumentation of LLM calls, tool usage, workflows, and evaluation signals (quality metrics, benchmarking, regression detection)
• Experience building cost monitoring, forecasting, and optimization systems for AI workloads
• Familiarity with cloud and infrastructure tooling (e.g., AWS, Azure, Kubernetes, Terraform)
• Experience with agentic infrastructure concepts such as MCP servers, hooks, skills, subagents, sandboxing, and persistent memory patterns
• Active engagement with the agentic engineering frontier, including emerging patterns (e.g., harness vs. model, review debt, feedback loops)
• Demonstrated agent-native development practices (iterating with agents using testing, verification, and feedback loops)
• Strong security awareness for autonomous systems, including sandboxing, prompt injection risks, credential exposure, and guardrails