Leega
Leega
201 – 500 Darbinieki
APIKonsultācijasMārketings
Leega ir Latīņamerikas tehnoloģiju konsultāciju uzņēmums, kas koncentrējas uz datu analītiku, mākoņrisinājumiem un digitālo pārveidi. Uzņēmuma pakalpojumi ietver lietojumprogrammu izstrādi, mašīnmācīšanos, risku pārvaldības analītiku un atbalstu organizācijām, kas modernizē savu datu vidi. Leega sadarbojas ar Google Cloud, AWS un Microsoft Azure, palīdzot uzņēmumiem uzlabot datu pārvaldību un ieviest mākoņtehnoloģijas. Uzņēmums sevi arī raksturo kā pirmo reģiona organizāciju, kas saņēmusi Google Cloud sertifikātu datu analītikas jomā.

Senior AWS Data Engineer

Build AWS data pipelines, agent infrastructure, and governance for generative AI at a financial institution. Work with Bedrock, multi-agent systems, sensitive data, and multi-account environments.

Apraksts

  • Build and maintain pipelines for Bedrock Knowledge Bases and agents, covering data ingestion, cleansing, chunking, and vectorization.
  • Establish a centralized source of truth for BI and agent knowledge bases.
  • Set data contracts, schemas, freshness standards, and SLAs for agent data sources.
  • Enable multi-agent orchestration with Strands Agents, Bedrock, and AgentCore.
  • Make internal data and systems available as reusable tools through MCP servers and AgentCore Gateway.
  • Design agent state management and shared memory while isolating contexts and clients.
  • Enable agent-to-agent communication with A2A, EventBridge, SQS, and Step Functions, using idempotency, timeouts, retries, and loop protection.
  • Enforce least-privilege access with dedicated identities, scoped permissions, and audit trails.
  • Design cross-account AWS access with IAM roles and Lake Formation.
  • Provision agents, tools, and guardrails as infrastructure as code using Terraform.
  • Track costs with tags, budgets, and metrics by department, project, model, and agent.
  • Maintain governance, privacy, LGPD compliance, and regional data residency.
  • Provide end-to-end observability with OpenTelemetry, AgentCore Observability, and CloudWatch.
  • Support resilience and disaster recovery with queues, multi-region designs, and controlled degradation.

Prasības

  • Demonstrated proficiency in Python and SQL.
  • Experience with S3, Glue, Athena, IAM, Lake Formation, and KMS.
  • Experience using Terraform and CI/CD.
  • Practical experience in multi-account AWS environments.
  • Knowledge of event-driven architecture and distributed systems, including queues, retries, idempotency, and consistency.
  • Experience building or operating at least one production LLM application, RAG system, or tool-enabled agent.
  • Familiarity with AWS FinOps practices, including tagging, Cost Explorer, and budgets.
  • Preferred experience or knowledge: Amazon Bedrock, Knowledge Bases, Guardrails, multi-agent collaboration, AgentCore, Strands, LangGraph, CrewAI, MCP servers, the A2A protocol, S3 Vectors, OpenSearch, Milvus, GraphRAG, MLOps, LLMOps, Data Mesh, hub-and-spoke architecture, AWS or FinOps certifications, and financial services or regulated environments.
  • Communicate clearly with data scientists, business stakeholders, and external partners.
  • Apply a strong security mindset and handle sensitive data rigorously.
  • Make decisions with explicit trade-offs.
  • Work autonomously to document and define responsibilities across teams.

Priekšrocības

  • Ongoing opportunities for professional growth.
  • Hybrid schedule with three days in the office and two days working from home.
  • Work in a dynamic, collaborative company.
  • Professional development and training supported through investment in employees.

Saistītās vakances

Nav atrasts neviens darbs

Mēģiniet pielāgot vai notīrīt filtrus, lai atrastu vairāk iespēju.