AI Engineer (Agents)

Wave Group; Madrid; On-site; EUR 90,000 - 110,000; Full time

Job summary

Wave Group in Madrid is seeking a senior backend engineer to build reliable AI-driven enterprise services. You will design and operate production-grade Python backends, with a focus on observability, performance, and scale across multi-agent systems.

You will own the evaluation layer, tracing, cost tracking, and regression testing for prompts and models, plus end-to-end work across async services, LLM pipelines, and voice components when relevant.

Qualifications

• Fluent English, ideally bilingual with Spanish.
• Strong Python backend engineering with high-throughput systems.
• Experience building evaluation pipelines for non-deterministic systems.
• Proficiency in observability, tracing, logging and cost tracking.
• Cloud deployment and containerisation (AWS/GCP, Docker).
• Git, CI/CD, testing and code reviews as standard practice.
• Startup pace, ownership and proactive mindset.
• Experience with STT/TTS or voice integrations is a plus.

Skills

Python backend

Asynchronous programming

Message queues

Observability

Tracing

Logging

Cost tracking

Model evaluation

LLM pipelines

Docker

Cloud experience

Git CI/CD

React (Frontend)

Twilio / telephony

React

Tools

Celery

Redis

AWS

GCP

Docker

Kubernetes

LLM tooling

Twilio

Job description

Salary: up to ~€100k (flex for strong candidates)

+ up to 35% bonus (~20% standard, rest based on performance)

Equity: very generous, early stage equity

Location: Madrid (preferred) or Barcelona (1-2 office days)

Company: autonomous AI agents for enterprise clients

About the company

This rapidly scaling start-up is building an AI-native platform that lets enterprise clients run high-volume operations through autonomous, conversational agents — voice, chat and messaging, deployed across dozens of countries in over 60 languages, for some of the largest employers in the world.

The company's core technical challenge is multi-agent orchestration that stays rock-solid at scale — reliable, secure and observable whether it's deployed for a healthcare provider in the US or a retailer in Latin America.

Founded by an experienced team with two prior successful exits, backed by a top-tier European VC, and already trusted by household-name enterprise customers.

About the role

You’ll build, operate and continuously improve the reliability of the agentic AI systems running in production for enterprise clients — focused on how agents behave, decide, fail, recover and scale under real operational load, not just on shipping new features.

You’ll own the observability and evaluation layer that keeps autonomous agents trustworthy at scale — tracing, logging, cost tracking across model calls, regression testing through prompt and model changes, and alerting on failures, hallucinations and degraded performance before customers ever see them.

You’ll also work end-to-end across the stack: production-grade async Python services, LLM/RAG pipelines, and the conversational voice layer where relevant — STT/TTS, telephony, real-time latency.

Must have requirements (most of the below):

• Professional fluency in English (and ideally Spanish)
• Strong Python backend engineering — async programming, high-throughput/message-queue systems (Celery + Redis or equivalent), production-grade service design
• Experience building evaluation pipelines for non-deterministic systems — accuracy/consistency testing, regression testing across prompt and model changes, synthetic test data generation
• Production observability instincts — tracing, logging, metrics, alerting on failures and degraded performance, cost tracking and optimisation across model calls
• Cloud deployment and containerisation experience (AWS/GCP, Docker)
• Git, testing, CI/CD, code review — clean, testable, well-documented code as standard practice
• Startup pace and mentality - high energy, ownership, proactiveness, hunger to innovate and succeed, comfortable operating with ambiguity
• Conversational/voice AI experience — STT/TTS pipelines, turn-taking, telephony integration (e.g. Twilio), latency-sensitive real-time systems
• Frontend experience (ideally React)
• Experience building enterprise or B2B products
• Model fine-tuning or model-selection/cost-optimisation experience