Melbourne, Australia · AI/ML leadership

Diego Marinho de Oliveira

Senior Engineering Manager / Principal AI

Production Agentic AI, Search & Recommender Platforms

Career experience: REA Group SEEK Catho

Diego Marinho de Oliveira

AI/ML leader with 10+ years shipping production systems across Oceania, Asia and South America—leading global data science and ML engineering teams from strategy to customer impact in agents, search, recommendations and multimodal AI.

Production AI scope

  • Agentic AI Production platform Orchestration, evaluation, guardrails and cost control
  • Search & ranking 150+ experiments Across search, ranking and personalisation
  • Recommenders 8+ markets Unified platform reach across Asia and ANZ
  • Leadership 12 people Across up to three AI/ML teams
Jump to a section

Career

Professional experience

A record of building teams and shipping measurable AI systems across property, employment and hiring marketplaces.

Current role

Senior Engineering Manager / Principal AI

Machine Learning · REA Group (opens in a new tab) · Melbourne, Australia

Global digital property business · home to realestate.com.au.

— Present

  • Built and led data science and machine learning engineering teams across production agentic AI, AI search and personalization—12 people across up to three teams.
  • Launched the RED brand’s first GenAI property search end to end in under one month, validating conversational discovery in production.
  • Built the production agent platform behind AI search and conversational experiences, including orchestration, evaluation, observability, guardrails and cost controls.
  • Delivered real-time multimodal computer vision attribute extraction in production, covering 45+ property attributes.
Agentic AIAI agentsAI search Multimodal AIEvaluationGovernance

AI / Data Scientist Manager; Senior Data Scientist; Data Scientist

SEEK (opens in a new tab) · Melbourne, Australia

Online employment marketplace across Australia, New Zealand and six Asian markets.

  • Built and led a global data science team of approximately 10 while hiring across multiple AI teams.
  • Led the recommender platform transformation across Asia and ANZ, producing 5–10x gains and a 12x peak uplift in one Asia market.
  • Delivered a unified recommender platform across 8+ markets and real-time recommendations for millions of users across homepage, email and push.
  • Launched multilingual personalized AI search for JobsDB in Hong Kong with a 14% improvement on a core metric.
Recommender systemsSearch ranking ExperimentationMulti-market platforms

Lead Data Scientist

Catho (opens in a new tab) · São Paulo, Brazil

Brazilian technology and online recruitment platform.

  • Built candidate recommendation systems that delivered more than 115% uplift on core marketplace metrics.

Earlier engineering and machine learning roles · Machine Learning Engineer, RBS Group / Appus; Machine Learning Specialist, Zunnit Technologies; Software Engineer, Vale Mining / Visagio; Software Analyst, TOTVS.

Selected work

AI systems shipped

Representative systems spanning AI agents, multimodal inference, information retrieval and recommendations.

Production agent platform

Orchestration → evaluation → safe operations

Built the shared platform behind customer-facing AI search and conversational experiences, standardizing agent orchestration, observability, evaluation, guardrails and cost governance.

AI agentsTool useObservability

GenAI property search

Production launch in under one month

Turned natural-language property intent into a production discovery experience, moving from concept to customer-facing validation in under one month.

REA GenAI property-search coverage (opens in a new tab)
Generative AIAI searchProduction

Unified recommender platform

Unified platform across 8+ markets

Replaced isolated market solutions with a reusable platform serving millions of users across real-time homepage, email and push recommendations in Asia and ANZ.

RecommendersRankingExperimentation

Multimodal attribute extraction

45+ property attributes in real time

Delivered production computer vision and multimodal transformer inference that extracted semantic property-photo attributes for downstream search and listing quality.

Multimodal AIComputer visionTransformers

Expertise

Leadership + technical depth

Hands-on technical judgement paired with team, product and operating leadership for production AI.

Agentic AI & GenAI

AI agents, multi-agent orchestration, tool calling, stateful workflows, memory, structured outputs, Retrieval-Augmented Generation (RAG), Model Context Protocol (MCP), embeddings and fine-tuning.

AI search & recommendations

Information retrieval, hybrid and semantic search, search relevance, ranking and reranking, learning-to-rank, sequential recommenders and large-scale personalization.

Evaluation, safety & MLOps

Agent evaluation, continuous evaluation, online experimentation, observability, tracing, guardrails, AI governance, latency and cost control across AWS, GCP and Vertex AI.

Teams & strategy

Organization design, hiring, mentoring, executive OKRs and cross-functional product strategy—translating AI opportunities into reliable systems, reusable platforms and measurable outcomes.

Research & recognition

Grounded in the field

Industry recognition, recommender-systems research and a computer science foundation.

Selected recognition

  • Winner, REA Hackdays — Hack it Forward Award for Community, Social & Environmental Good (2024)
  • Top performer, Databricks Gen-AI Australia Cup (2023)

Languages: English (fluent), Portuguese (native), Spanish (basic).

Let’s discuss production AI leadership.

For senior AI/ML leadership, agentic AI platform, search or recommendation opportunities, email me or connect on LinkedIn.