AI Engineer - Azure & Agentic Systems (Brazil)

Elios, Inc.; Brasil; Teletrabalho; BRL 120 000 - 190 000; Tempo integral

Vantagens oferecidas por esta oferta de emprego

Remote work from Brazil

Resumo da oferta

Elios, Inc. is seeking an experienced engineer to design and build agentic automations on Azure, combining low-code workflows with Python.

You will work remotely from Brazil on a contract basis, handling triggers, connectors, and control flow with Azure Logic Apps (Standard) and custom Python logic. The role requires 3–5+ years in software engineering, strong Python and SQL skills, and experience with Azure services, including Key Vault and Entra ID.

Qualificações

• 3–5+ years of professional software engineering experience with Python as a primary language.
• Hands-on experience across the Azure ecosystem including identity and secrets management.
• Experience with REST APIs: consuming and exposing endpoints.
• Strong English communication for collaboration with US-based team.

Responsabilidades

• Build agentic orchestrations using Azure Logic Apps (Standard) and Azure AI Foundry Agent Service.
• Design stateful and stateless workflows with connectors and triggers.
• Write Python for integrations and custom logic beyond low-code capabilities.
• Expose and document REST APIs, and consume external endpoints.
• Integrate databases into automated workflows with solid SQL.
• Implement identity and secrets management with managed identities and Key Vault.
• Develop prompt/context engineering and evaluation methods for agent outputs.
• Advise on when agentic vs deterministic approaches are appropriate.

Conhecimentos

Python

SQL

REST APIs

Azure

English

LLM agents

Ferramentas

Azure Logic Apps

Azure AI Foundry

Azure Key Vault

Entra ID

Descrição da oferta de emprego

Workplace Type Remote

Position Type Contract

Sector Technology

Years of Experience 3+ years

Job Description

Location: Remote from Brazil | Type: Contract, full-time hours

About the Role

You'll design and build agentic orchestrations on Azure, combining low-code workflow design with custom Python where the logic gets interesting. In practice, that means Azure Logic Apps (Standard) handling triggers, connectors, and control flow, with Python carrying the work no connector can do.

This is a broad Azure role, not a narrow one. In the same week, you might configure managed identities and pull secrets from Key Vault, integrate a SQL database into a workflow, expose a REST endpoint, then tune prompts and build an evaluation harness for agent output.

The company ships production systems for large enterprises, so the agents you build need to hold up against real users and real data, not just demo cleanly.

You should also have a clear point of view on where an agent belongs and where a deterministic workflow is the better answer. Not every problem needs autonomous multi-step planning, and the person in this seat should be able to say so and propose the simpler design.

Our client operates with end-to-end ownership, short delivery cycles, and a high bar for craft. Self-starters do well here. People who need problems pre-defined usually don't.

Responsibilities

• Build agentic orchestrations using Azure Logic Apps (Standard) and Azure AI Foundry Agent Service, mixing low-code workflow design with custom Python
• Design stateful and stateless workflows, configure connectors and triggers, and handle control flow that survives messy real-world inputs
• Write Python for integrations, internal tooling, and the custom logic low-code can't cover
• You'll consume external REST APIs and also design, expose, and document your own
• Integrate databases into automated workflows using solid SQL
• Implement identity and secrets management correctly with managed identities, Entra ID, and Azure Key Vault
• Handle prompt and context engineering, then build evaluations so agent output quality is measured rather than assumed
• Make and defend architecture calls on agentic versus deterministic approaches

Qualifications

• Hands-on experience across the Azure ecosystem: compute, storage, networking, identity (managed identities / Entra ID), and secrets (Key Vault)
• Azure Logic Apps (Standard): connectors, triggers, stateful and stateless workflows, and control flow
• Azure AI Foundry Agent Service experience building, configuring, and consuming agents, or directly transferable experience with LLM agents and tool use
• 3 to 5+ years of professional software engineering experience with Python as a primary language
• Strong SQL skills and real comfort integrating databases into workflows
• Experience on both sides of REST APIs: consuming external endpoints and designing and exposing your own
• Working knowledge of agentic patterns: tool use, orchestration, prompt and context engineering, and evaluation of agent outputs
• Written and spoken English strong enough for direct collaboration with a US-based team

Preferred Skills

• Consulting or agency background, or other experience delivering directly to enterprise stakeholders
• Already using AI tooling in day-to-day workflow rather than just curious about it
• Experience with observability, cost, and latency management on LLM-backed systems
• Familiarity with Azure Functions, Service Bus, or other event-driven Azure services

Tech Stack (Representative)

• Languages: Python, SQL
• Integration: REST APIs, connectors, webhooks, JSON
• AI: LLM agents, tool use, prompt and context engineering, agent evaluation

Closing

This fits an engineer who is comfortable owning the whole path from workflow diagram to working integration, who knows Azure well enough to work without a platform team guiding every step, and who has the judgment to push back when an agent is the wrong tool for the job.