Principal Data & AI Architect

This is Us

Ingenious Lion is a Digital Tech Hub designed exclusively for Publicis Groupe, one of the world's leading marketing agencies.

As the marketing industry continues to drive digital transformation, Publicis Groupe empowers its clients with innovative data solutions and media ecosystems. Ingenious Lion has one mission: to partner with Publicis' global teams in creating innovative data and digital tech products that will shape the future.

What makes us successful? Two key ingredients: we seek the good and the best, meaning we cultivate a thriving company culture. We're all about creating an environment where people don't just work; they grow, challenge themselves, strive for excellence, and are recognized for their incredible value to the team.

If you're curious to learn more about us, here are four things you should know:

• Our company culture is anchored in five powerful values: Joy, Honesty, Connectivity, Own Challenger, and Greater World. Do these values resonate with you?

• We are committed to diversity and inclusion because we believe our unique differences make us stronger.

• We encourage open communication and create an environment where everyone's voice is heard and every contribution is valued.

• Finally, we’re somewhat “obsessed” with quality. We believe in continuous learning and knowledge sharing; when we hit a roadblock, we don’t just complain, we create solutions to overcome it.

In one sentence, we are a group of people with an unusual aptitude for generating great ideas, groundbreaking concepts, and infusing them with positive energy and good fun.

Feeling Ingenious already?

A Sneak Peek at Your Future Role

As a Principal Data & AI Architect, you will set the technical direction for OneSuite data products and their integration into the Performics platform. This is a hands-on principal role: you will write code, design cross-repo architecture, lead high-stakes technical decisions, and raise the standards that other engineers use to build reliable, scalable systems. This is a senior individual-contributor / specialist-track position — the highest level of technical expertise on this track — you'll set technical direction and mentor other engineers, but it does not include people-management responsibilities.

You will help build the backend and data foundation behind AI-assisted analyst workflows, including agents built with the Claude Agent SDK, turning complex marketing data into dependable services, tools, and agent experiences that can answer questions, surface issues, and support campaign decisions securely across clients.

Working closely with product and data teams, you will own architecture across marketing data pipelines, connector systems, platform APIs, and agent/tool runtimes, ensuring systems are scalable, reliable, secure, and built to evolve.

This role is for someone who already thinks natively in agentic architecture. You'll bring that depth with you from day one and put it straight to work — becoming the team's go-to voice on AI and agent systems, while ramping up on the business and day-to-day context as you go.

Here Are the Key Tasks You'll Be Working on Every Day

• Own backend and data architecture across marketing data pipelines, connector systems, platform APIs, agent/tool runtimes, and downstream product surfaces, defining durable domain boundaries and data contracts between warehouse outputs, platform APIs, agent tools, and user-facing workflows.

• Turn ambiguous platform needs into incremental technical plans that balance reliability, security, tenant isolation, developer experience, and long-term maintainability.

• Lead the evolution of data pipelines following the medallion architecture, driving data quality, schema enforcement, lineage, and curated outputs for analytics and product use cases.

• Harden ingestion from external marketing platforms and related APIs, including connector and reconciliation patterns.

• Architect backend services across API paths and cloud infrastructure, shaping how platform capabilities are exposed through stable contracts, streaming workflows, and tool permissions.

• Design and own the architecture of agent and tool-runtime systems, including tool permissioning, secure multi-tenant agent access, and durable boundaries between agent tools and platform APIs.

• Implement secure authentication, authorization, rate limiting, and multi-tenant access controls across services.

• Write high-quality, production-ready code and integrate with external APIs, cloud services, internal microservices, and agent/tool runtimes.

• Build and maintain deployment pipelines and infrastructure environments, implementing tracing, structured logging, metrics, and alerting to ensure reliability, latency, and operational standards.

• Mentor engineers — including on agentic system design and AI integration — contribute to internal libraries and backend frameworks, and drive consistency across the platform.

We're Thrilled to Have You Consider Joining Our Team. Here Are the Qualifications We're Seeking

• At least ten years of experience in backend, data platform, or distributed systems, including ownership across multiple services or teams.

• At least two years of hands-on, production experience building with LLMs and agent frameworks — not prototypes, coursework, or exploratory side projects — including writing and shipping code that integrates agent/tool runtimes (Claude Agent SDK or MCP-style tooling) into live systems.

• Demonstrated experience architecting agent and tool-runtime systems: defining tool permissioning, durable domain boundaries between agent tools and platform APIs, and secure multi-tenant access patterns for agentic workflows — beyond simply calling an LLM API from within an existing service.

• Expertise in Python, API design, service boundaries, async and concurrent systems, and modern testing practices.

• Proficiency with Databricks, PySpark, Lakehouse patterns, and medallion-style data architecture — or equivalent large-scale data pipeline experience with strong enough architectural fundamentals to pick up medallion patterns quickly, since that's how our pipelines are structured today.

• Strong AWS serverless experience, including Lambda, API Gateway, S3, IAM, and CloudFormation/SAM or equivalent infrastructure-as-code tooling.

• Deep understanding of multi-tenant SaaS design, authentication and authorization, tenant-scoped data access, secrets management, and operational security.

• Experience designing stable data contracts, schema evolution patterns, and consumer-facing data models.

• Proficiency with SQL and NoSQL databases such as PostgreSQL, Redis, or MongoDB.

• Strong CI/CD experience and deep knowledge of observability practices including logs, metrics, and traces.

• Technical leadership through mentoring, architecture decisions, code reviews, communication, and engineering standards — including the ability to bring engineers who are new to agentic systems up to speed.

• Fluent in English, both verbal and written.

Skills That Would Make You Shine Even Brighter!

• You have direct contributions to the Claude Agent SDK, MCP tooling, or similar agent-framework ecosystems.

• You are familiar with Fivetran, custom connector development, or third-party data ingestion operations.

• You have had contact with Delta Lake, Unity Catalog, data governance, data lineage, and Lakehouse performance optimization.

• You have knowledge of streaming or queuing systems such as Kafka, Pub/Sub, or SQS.

• You are familiar with marketing, advertising, or platform APIs and similar ecosystems.