AI Analytic Developer- Federal Projects

Oracle Corporation; United States; Presencial; USD 180.000 - 240.000; Jornada completa

Descripción de la vacante

Oracle Health is seeking a Senior AI Agent Engineer to build production AI agents and workflow automation capabilities that accelerate analytics delivery, improve insight generation, and reduce manual engineering work across our analytics and reporting ecosystem.

This engineer will design and implement AI-driven workflows that integrate with enterprise systems and support use cases such as SQL generation, pipeline scaffolding, data exploration, metric design support, insight discovery, executive

Formación

• 8+ years of software engineering experience, including hands-on LLM-enabled or agentic AI solutions.
• Experience with AI agents and workflow automation, including tool calls and orchestration.
• Expertise in enterprise GenAI, retrieval architectures, and secure production systems.

Responsabilidades

• Design, build, test, and deploy production AI agents and AI-assisted workflows for analytics and reporting.
• Implement agents that support tasks such as SQL generation, pipeline creation, metric design support, insight discovery, executive narrative generation, and dashboard/report scaffolding.
• Integrate AI capabilities with enterprise tools, data platforms, semantic services, and existing engineering workflows.
• Develop structured tool‑calling patterns, validation logic, retry handling, and grounded output generation.
• Build evaluation and testing approaches for AI workflows, including regression testing, quality validation, and failure analysis.

Conocimientos

Python

APIs

Distributed systems

LLM-enabled applications

Agent orchestration

RAG

Observability

Herramientas

OCI Generative AI

Vector search

OpenAI API

Descripción del empleo

Oracle Health is seeking a Senior AI Agent Engineer to build production AI agents and workflow automation capabilities that accelerate analytics delivery, improve insight generation, and reduce manual engineering work across our analytics and reporting ecosystem.

This engineer will design and implement AI-driven workflows that integrate with enterprise systems and support use cases such as SQL generation, pipeline scaffolding, data exploration, metric design support, insight discovery, executive narrative creation, and report/dashboard acceleration. The role requires strong software engineering fundamentals combined with hands‑on experience building LLM-enabled applications that are grounded, testable, and production-ready.

The ideal candidate is a builder who can move from concept to implementation, balancing speed with safety, evaluation, and maintainability. This person should be able to work across prompts, tools, code, APIs, structured outputs, validation, and workflow integration to deliver real productivity gains.

Requirements

• 8+ years of software engineering experience, with strong proficiency in Python, APIs, distributed systems, and production-grade application development, including hands‑on experience building and deploying LLM-enabled or agentic AI solutions.
• Demonstrated expertise in AI agents and workflow automation, including tool/function calling, agent orchestration, prompt and context management, structured outputs, validation, retry/failure handling, human‑in‑the‑loop workflows, and integration with enterprise systems.
• Strong experience with enterprise GenAI and retrieval architectures, including RAG, embeddings, vector and hybrid search, reranking, grounding, knowledge bases, semantic layers, and NL2SQL/conversational analytics; experience with OCI Generative AI, Oracle AI Database/Vector Search, Select AI, OpenSearch, or comparable technologies preferred.
• Proven ability to build production-ready, secure, and measurable AI systems, including evaluation frameworks, regression testing, hallucination and grounding assessment, observability, latency/cost optimization, access controls, auditability, and safeguards against prompt injection and data exfiltration.
• Strong architecture and cross‑functional collaboration skills, with the ability to define reusable APIs, schemas, tool interfaces, integration patterns, and engineering standards while partnering with data, analytics, security, application, infrastructure, and domain teams to translate complex business requirements into scalable AI solutions.

Internal Responsibilities

• Design, build, test, and deploy production AI agents and AI‑assisted workflows for analytics and reporting use cases.
• Implement agents that support tasks such as SQL generation, pipeline creation, metric design support, insight discovery, executive narrative generation, and dashboard/report scaffolding.
• Integrate AI capabilities with enterprise tools, data platforms, semantic services, and existing engineering workflows.
• Develop structured tool‑calling patterns, validation logic, retry handling, and grounded output generation.
• Build evaluation and testing approaches for AI workflows, including regression testing, quality validation, and failure analysis.
• Partner with architects, semantic engineers, and domain teams to ensure AI outputs are useful, accurate, and operationally practical.
• Optimize workflows for quality, speed, cost, and maintainability.
• Contribute to AI engineering standards, reusable components, and continuous improvement across the team.
• Foundation‑model access, model routing, inference, embeddings, reranking, fine‑tuning, prompt/context management, and model lifecycle.
• Agent runtimes, multi‑agent orchestration, planning, memory, tool execution, workflow integration, human approvals, and secure delegation.
• Retrieval‑augmented generation (RAG), knowledge bases, semantic and hybrid search, query rewriting, reranking, grounding, provenance, and citations.
• Enterprise knowledge management including ingestion, parsing, chunking, metadata, taxonomy, ontology, business glossary, lifecycle, and entitlement‑aware retrieval.
• NL2SQL and conversational analytics, including schema/semantic grounding, metadata enrichment, SQL generation and validation, permission‑aware execution, and natural‑language narration of results.
• AI‑ready data platforms supporting batch, streaming, change data capture, lakehouse patterns, data products, feature/embedding generation, and low‑latency serving.
• AI evaluation and observability covering quality, hallucination/grounding, retrieval relevance, tool‑call success, agent completion, latency, reliability, safety, and cost.
• AI security and governance including identity, authorization, tenant isolation, private networking, secrets, auditability, data residency, prompt‑injection defenses, exfiltration controls, and policy enforcement.
• Multimodal AI experiences spanning text, documents, images, speech, and structured enterprise data.
• Developer platforms, APIs, SDKs, reference implementations, and reusable components that let Oracle teams and customers build AI applications consistently.
• Define architecture patterns that combine OCI Generative AI and agent capabilities with Oracle AI Database 26ai, Autonomous AI Database, AI Vector Search, Select AI, Oracle AI Data Platform, GoldenGate, OpenSearch, Object Storage, and other OCI data services.
• Establish reusable integration patterns between OCI AI services and Oracle Fusion Cloud Applications, AI Agent Studio / Fusion Agentic Applications, Oracle Analytics Cloud, Fusion Data Intelligence, Oracle Integration, and adjacent Oracle products.
• Drive semantic architecture across structured data and enterprise knowledge so agents and AI assistants understand business concepts, relationships, permissions, and source‑of‑truth boundaries rather than only raw schemas or documents.
• Define retrieval architecture choices across Oracle AI Database vector search, OCI Search with OpenSearch, managed knowledge bases, file search, and federated enterprise sources, including guidance for hybrid retrieval, ranking, freshness, and ACL enforcement.
• Partner with database, data, AI science, applications, analytics, security, and infrastructure teams to operationalize new model capabilities without fragmenting the platform architecture.
• Define canonical APIs, schemas, contracts, tool interfaces, event patterns, and interoperability standards for agents, models, knowledge sources, data products, and enterprise applications.
• Establish patterns for hybrid and distributed deployments where data or inference must remain close to regulated, sovereign, customer, or on‑premises environments.
• Evaluate emerging AI, data, search, agent, and model‑serving technologies and determine where Oracle should build, integrate, standardize, or partner.

External Responsibilities

• Design, build, test, and deploy production AI agents and AI‑assisted workflows for analytics and reporting use cases.
• Implement agents that support tasks such as SQL generation, pipeline creation, metric design support, insight discovery, executive narrative generation, and dashboard/report scaffolding.
• Integrate AI capabilities with enterprise tools, data platforms, semantic services, and existing engineering workflows.
• Develop structured tool‑calling patterns, validation logic, retry handling, and grounded output generation.
• Build evaluation and testing approaches for AI workflows, including regression testing, quality validation, and failure analysis.
• Partner with architects, semantic engineers, and domain teams to ensure AI outputs are useful, accurate, and operationally practical.
• Optimize workflows for quality, speed, cost, and maintainability.
• Contribute to AI engineering standards, reusable components, and continuous improvement across the team.
• Foundation‑model access, model routing, inference, embeddings, reranking, fine‑tuning, prompt/context management, and model lifecycle.
• Agent runtimes, multi‑agent orchestration, planning, memory, tool execution, workflow integration, human approvals, and secure delegation.
• Retrieval‑augmented generation (RAG), knowledge bases, semantic and hybrid search, query rewriting, reranking, grounding, provenance, and citations.
• Enterprise knowledge management including ingestion, parsing, chunking, metadata, taxonomy, ontology, business glossary, lifecycle, and entitlement‑aware retrieval.
• NL2SQL and conversational analytics, including schema/semantic grounding, metadata enrichment, SQL generation and validation, permission‑aware execution, and natural‑language narration of results.
• AI‑ready data platforms supporting batch, streaming, change data capture, lakehouse patterns, data products, feature/embedding generation, and low‑latency serving.
• AI evaluation and observability covering quality, hallucination/grounding, retrieval relevance, tool‑call success, agent completion, latency, reliability, safety, and cost.
• AI security and governance including identity, authorization, tenant isolation, private networking, secrets, auditability, data residency, prompt‑injection defenses, exfiltration controls, and policy enforcement.
• Multimodal AI experiences spanning text, documents, images, speech, and structured enterprise data.
• Developer platforms, APIs, SDKs, reference implementations, and reusable components that let Oracle teams and customers build AI applications consistently.
• Define architecture patterns that combine OCI Generative AI and agent capabilities with Oracle AI Database 26ai, Autonomous AI Database, AI Vector Search, Select AI, Oracle AI Data Platform, GoldenGate, OpenSearch, Object Storage, and other OCI data services.
• Establish reusable integration patterns between OCI AI services and Oracle Fusion Cloud Applications, AI Agent Studio / Fusion Agentic Applications, Oracle Analytics Cloud, Fusion Data Intelligence, Oracle Integration, and adjacent Oracle products.
• Drive semantic architecture across structured data and enterprise knowledge so agents and AI assistants understand business concepts, relationships, permissions, and source‑of‑truth boundaries rather than only raw schemas or documents.
• Define retrieval architecture choices across Oracle AI Database vector search, OCI Search with OpenSearch, managed knowledge bases, file search, and federated enterprise sources, including guidance for hybrid retrieval, ranking, freshness, and ACL enforcement.
• Partner with database, data, AI science, applications, analytics, security, and infrastructure teams to operationalize new model capabilities without fragmenting the platform architecture.
• Define canonical APIs, schemas, contracts, tool interfaces, event patterns, and interoperability standards for agents, models, knowledge sources, data products, and enterprise applications.
• Establish patterns for hybrid and distributed deployments where data or inference must remain close to regulated, sovereign, customer, or on‑premises environments.
• Evaluate emerging AI, data, search, agent, and model‑serving technologies and determine where Oracle should build, integrate, standardize, or partner.