Эта вакансия в архиве
Посмотреть похожие вакансии ↓10 дней назад
GraphRAG Engineer (AI)
Описание вакансии
Текст:
TL;DR
GraphRAG Engineer (AI): Engineering the semantic, vector, and graph storage layer for an enterprise AI-first platform with an accent on Neo4j, pgvector, distributed event streaming, and multi-cloud infrastructure. Focus on building SDLC context graphs and hybrid retrieval pipelines, constraining autonomous agents, and ensuring high availability across AWS and GCP.
Location: Budapest, Hungary; hybrid work
Company
provides enterprise data management and analytics solutions that help organizations transform complex data into actionable insights.
What you will do
- Provision, tune, and maintain production-grade Neo4j graph database and pgvector storage clusters.
- Build ingestion pipelines for Git repositories, ASTs, Jira links, Avro schemas, and CI/CD metadata to create an SDLC Context Graph.
- Orchestrate hybrid GraphRAG retrieval using SQL, Cypher graph traversals, and dense vector embeddings.
- Configure safeguards such as circuit breakers, confidence thresholds, and execution limits for autonomous agents.
- Integrate knowledge stores and microservices with the Enterprise AI Gateway, including version-controlled prompts and DLP/PII controls.
- Implement failover, backup restoration, and multi-cloud storage cost controls across AWS and GCP.
Requirements
- Deep operational and development experience with Neo4j, Cypher, APOC, causal clustering, or enterprise knowledge graphs.
- Proven experience with pgvector, PostgreSQL, embeddings management, hybrid search, and RAG frameworks or custom pipelines.
- Hands-on database administration across AWS and GCP cloud environments.
- Experience consuming Apache Avro payloads, streaming Kafka events through AWS MSK, and parsing code and JSON artifacts.
- Practical knowledge of Prompts-as-Code, few-shot prompt optimization, and agent tool specification.
Nice to have
- Infrastructure-as-Code experience with Terraform, Kubernetes, Docker, and pull-based GitOps workflows.
- Experience with HashiCorp Vault Transit encryption, OIDC keyless authentication, and zero-trust workload identities.
- Familiarity with OpenTelemetry instrumentation and Datadog or Grafana for monitoring vector search and LLM performance.
Culture & Benefits
- Flexible work-from-home policy within the hybrid setup.
- Generous paid time off and designated unplugged days.
- Mental and physical wellness programs.
- Phone and internet reimbursement.
- Career development opportunities, comprehensive benefits, paid volunteer time, and employee resource groups.