Python Developer (ai)

Box Box · Enterprise · Poland · Remote · IT

Develop and deploy AI agents and operational solutions to improve IT and business workflows, focusing on minimizing hallucinations, prioritizing security, and enabling informed decision-making. This role involves building and maintaining infrastructure, data pipelines, and developer tools for AI capabilities.

What you'd actually do

  1. Design, build, and deploy AI agents that automate and improve IT and business workflows while minimizing hallucinations and ensuring reliable outputs.
  2. Implement model validation, monitoring, and fail-safe mechanisms to detect and reduce incorrect or risky agent behavior.
  3. Architect secure data handling pipelines (ingest, storage, access controls, encryption) to protect sensitive information used by models and agents.
  4. Develop and maintain MCP servers and developer tooling that surface actionable insights and make it easy for teams to query, debug, and act on model outputs.
  5. Integrate observability and telemetry (logging, metrics, tracing) for agents and infrastructure to enable rapid incident detection, root-cause analysis, and performance tuning.

Skills

Required

  • Python
  • LLMs
  • prompt engineering
  • RAG
  • vector stores
  • ML frameworks (PyTorch/TensorFlow)
  • model evaluation/validation techniques
  • Cloud experience (AWS/Azure/GCP)
  • Kubernetes
  • Docker
  • Terraform
  • Observability and monitoring tools (Prometheus, Grafana, ELK, OpenTelemetry)
  • Security and compliance familiarity (IAM, encryption, GDPR/HIPAA awareness)
  • Cloud data warehouses (BigQuery, Snowflake, Redshift, or similar)
  • lead cross-functional initiatives
  • communicate technical concepts

Nice to have

  • Slack app development
  • Google Cloud Platform
  • BigQuery administration
  • data access/security modeling
  • LangChain
  • Ray
  • agent frameworks
  • semantic metadata/catalog tooling (Dataplex, Looker)
  • CI/CD (GitHub Actions/Jenkins)
  • chaos testing
  • load testing

What the JD emphasized

  • minimize hallucinations
  • prioritize security
  • secure data handling pipelines
  • low-risk decision-making
  • minimize hallucinations
  • reliable outputs
  • incorrect or risky agent behavior
  • protect sensitive information
  • safely interact
  • secure, low-risk usage
  • security, compliance
  • secure, scalable AI platform
  • trusted decisions
  • secure data sharing
  • regulated data classification experience
  • reducing hallucinations
  • improving model reliability

Other signals

  • design and deploy AI agents
  • minimize hallucinations
  • secure data handling pipelines
  • integrate observability and telemetry
  • optimize infrastructure cost and reliability