About this role
Data Scientist – Agentic AI
Permanent
Location Greece
LIFE AS A SATALIAN
At Satalia, a WPP company, we push the boundaries of data science, optimisation and AI to solve the most complex problems in the industry. Led by our founder and WPP Chief AI Officer Daniel Hulme, Satalia’s ambition is to become a decentralised organisation of the future, developing tools and processes to liberate and automate manual repetitive tasks, with a focus on freedom, transparency and trust.
We are a community working on diverse and challenging projects, where you can flex your technical skills whilst working alongside high performing colleagues. We offer truly flexible working, prioritise wellbeing and inclusivity, and create a safe environment for innovation and growth, where your opinion matters and achievements are celebrated.
THE ROLE
We are at a generational moment for technology and AI. The shift from single-model applications to autonomous, collaborative agent systems represents the most significant leap since the introduction of Large Language Models (LLMs). At Satalia, we are not just observing this transformation; we are building it. These are not experimental prototypes; they are production-grade solutions that make a tangible impact for our FTSE 100 clients and WPP agency partners, at enterprise scale.
If you have been waiting for the right opportunity to go deep on agentic AI, this is it.
WHAT YOU WILL BE DOING
Design and deliver production-grade agentic systems, from multi-agent orchestration to the AI services and models that power them.
Assess fit of agentic solutions per use case, being intentional in system design
Design the most suitable orchestration pattern per use case
Evaluate and recommend the best suited agent development framework
Engineer multi-step agent workflows that are reliable, auditable and modular, supporting both linear and non-linear control flows
Implement agent memory and state management systems
Optimise context assembly for each agent interaction
Evaluate and select the most appropriate foundation models, embeddings and context-specific model variants; apply fine-tuning or adaptation where needed
Implement guardrails to ensure operation within ethical, legal, brand boundaries
Build services for agent use, including ML models, GenAI and task-specific APIs
Process multi-modal data using appropriate embeddings and vector retrieval
Build, evaluate and maintain robust RAG and / or GraphRAG pipelines
Implement fallback strategies (e.g., retries, backup tools, escalation, safe exits)
Build and integrate MCP servers to expose and consume data and AI services
Use Google's A2A protocol to support independent agent communication, designing composable service boundaries and appropriate integration patterns
Containerise applications (Docker) – our platform team manages K8s
Apply AgenticOps practices, including evaluation, observability, guardrails, security, performance optimisation, CI/CD, versioning, rollback, drift detection
Design and implement testing strategies for non-deterministic behaviour
Stay abreast of the latest AI research trends and integrate them into our products
Maintain documentation for architectures, APIs and operational procedures
Communicate progress, blockers, technical approaches, challenges and potential solutions to technical colleagues and non-technical stakeholders.
WHAT WE ARE LOOKING FOR
Education & Experience
Educational background in Computer Science, Data Science, or similar
Professional experience in Data Science, AI/ML Engineering, Software Engineering, or related technical roles
Core Programming & Development
High proficiency in Python and working knowledge of SQL
Experience building API services using FastAPI, Flask or similar
Experience with containerisation, cloud platforms and CI/CD
Experience with AgenticOps best practices, including evaluation, observability, safety guardrails, security, performance optimisation
Agentic AI
Experience designing and building complex multi-agent systems using at least one of the major agent frameworks (e.g. ADK, AutoGen, LangGraph)
Understanding of agent memory and state management patterns
Familiarity with MCP servers
Quality-first, test-driven mindset with focus on automated testing
AI/ML & Data
Deep understanding of LLMs, foundation models and embeddings
Experience with structured and multi-modal data processing
Experience building and evaluating production-grade RAG systems
Practical experience with machine learning / deep learning models
Communication & Collaboration
Excellent communication skills with technical / non-technical audiences
Strong documentation and knowledge-sharing practices.
EVEN BETTER IF YOU HAVE:
Postgraduate degree in Computer Science, Data Science or similar
Experience with Google's Agent Development Kit (ADK)
Experience with Google's A2A protocol and agent coordination layers
Strong understanding of ML / DL algorithms and full model lifecycle experience
Experience with fine-tuning or model adaptation techniques
Experience with GraphRAG systems
Understanding of authentication/authorisation and API security
K8s familiarity for cloud-native design and debugging. Our Platform team manages the clusters.
Satalia is home to some of the brightest minds in AI and if you’re looking to join a company who not only values autonomy and freedom, but embraces a culture of inclusion and warmth, we’d love to hear from you.
We aim to respond to all applications within 2 weeks. If you have not heard from us within 2 weeks this means your application has been unsuccessful. By applying to Satalia you are expressly giving your consent for the collection and use of your information as described within our Satalia Recruitment Privacy Policy. Good luck!