Privado.ai
Website:
privado.ai
Job details:
You will build interactive AI agents that execute privacy auditing tasks across websites and iOS/Android apps. These agents will plan and run audits, integrate with tools, collect evidence, and produce actionable compliance outputs reliably and at scale.
Responsibilities
- Build task-completing agents that run end-to-end audits: plan, execute, observe, iterate, and finish with clear completion criteria and safe fallbacks.
- Implement consent verification workflows (accept, reject, no action) and validate how tracking behavior changes across configurations and locations.
- Integrate agents with tools and services to enable audit execution (scan runners, traffic inspection, evidence capture, and reporting pipelines).
- Implement stateful workflows with task tracking and durable state so audits can be resumed, compared across releases, and run continuously.
- Create evaluation and observability loops: scenario-based tests, regressions, tracing/logging, and production monitoring to improve reliability and completion rates.
- Collaborate with Product/Design to ship an interactive experience with progress, explanations, confirmations, and actionable outputs.
Requirements
- 1-3 years of software engineering experience; strong Python fundamentals.
- Hands-on experience building interactive AI agents that complete multi-step tasks (not just chat): tool use, workflow orchestration, and state/memory management.
- Experience integrating with APIs and services and building backend endpoints for long-running workflows.
- Experience testing and debugging agent behavior using evaluations and runtime observability.
- Understanding of data structures, algorithms, and software design principles.
- Familiarity with production delivery (Docker, CI/CD, cloud basics).
Tech Stack (what You Will Use)
- Agent orchestration: LangGraph/LangChain (plus exposure to LlamaIndex, AutoGen, CrewAI, Semantic Kernel, Haystack, or similar).
- Backend: Python (Flask/Django).
- Frontend (helpful): React.
- Integrations/services (helpful): Node.js .
- Data stores: MongoDB (plus durable workflow/state storage as needed).
Nice To Have
- Multi-agent patterns (parallel specialists, coordination, routing).
- Experience building evidence-grade reports (reproducible outputs, clear pass/fail logic, traceability).
- Interest or exposure to privacy/compliance workflows (consent validation, tracking governance, third-party data flow visibility).
- Open-source contributions to agent tooling or developer platforms.
This job was posted by Prashant Mahajan from privado.ai.
Click on Apply to know more.