Course Objective & Scope
Traces and logs enabled for debugging and quality monitoring in Langfuse, LangSmith, Arize Phoenix or application loggers often become a shadow repository storing raw personal data, financial details and confidential RAG chunks in cleartext. This course shows how to achieve data minimisation and KVKK compliance in logging pipelines without sacrificing operational observability or security auditing.
Participants build OpenTelemetry Collector scrubbing processors, tiered logging (metadata-only vs payload capture), encrypted vault storage, automated TTL retention and data-subject deletion workflows.
Who Should Attend?
- LLMOps, DevOps, SRE and platform engineers
- Teams managing observability and central logging infrastructure
- Software and security engineers working alongside data protection and KVKK compliance teams
- Prerequisite: familiarity with application logging, OpenTelemetry or LLMOps tracing tools
Key Takeaways
- Audit existing LLM traces, logs and evaluation stores for KVKK and privacy exposures
- Implement pre-ingestion PII scrubbing filters in OpenTelemetry and LLMOps SDKs
- Design zero-content telemetry schemas capturing tokens, latency, cost and safety scores without raw text
- Isolate full-payload debug captures behind separate encryption keys, strict RBAC and short TTLs
- Codify retention and disposal policies into a technical checklist ready for KVKK audits
Curriculum
The Risk of Hidden Data Accumulation in LLM Telemetry
- Data leaked in prompts, completions, RAG chunks, tool arguments and error stack traces
- Cross-border transfer and KVKK obligations when using third-party SaaS observability tools
Data-Minimised Telemetry Schema Design
- Metadata-only metrics: token counts, latency, model ID, guardrail scores and content hashes
- Conditional and sampled payload capture: recording raw text only on explicit consent or policy alerts
Pre-Ingestion PII Scrubbing Pipelines
- OpenTelemetry Collector redaction/transform processors and SDK callback filters
- Recording redacted traces in self-hosted Langfuse or Arize Phoenix deployments
Access Control, Encryption and Automated Retention / Disposal
- Separating operational telemetry from an encrypted payload vault with strict RBAC
- Automated index TTLs, cold storage tiering and cryptographic erasure (crypto-shredding)
KVKK Audit Evidence and Data Subject Requests
- Aligning logging architecture with the data inventory (VERBİS), privacy notices and retention policies
- Purging user records across traces, semantic caches and evaluation datasets upon deletion requests
Other Courses in This Category
-
Covers Agent Goal Hijack—the top OWASP Agentic AI risk—showing how attackers divert autonomous planning loops, persistent memory and multi-step execution, and how to defend against it with intent verification.
- Live Course
- 5 modules
-
Covers tool poisoning, rug pulls, tool shadowing and supply chain risks in Model Context Protocol (MCP) servers, third-party agent tools and A2A integrations, and how to build a secure MCP architecture.
- Live Course
- 5 modules
-
Teaches how to prevent remote code execution (RCE) and injection vulnerabilities caused by executing LLM- and agent-generated Python, SQL, shell or template outputs using AST validation and safe output handling.
- Live Course
- 5 modules
-
Teaches how to isolate code-executing, file-processing and browser-using AI agents inside secure sandboxes using Firecracker microVMs, gVisor, Kata Containers, WebAssembly and strict network egress controls.
- Live Course
- 5 modules
-
Hands-on training on direct prompt injection, jailbreak techniques and system prompt leakage, showing developers how to detect and defend against them with input classifiers, instruction hierarchy and tools such as garak.
- Live Course
- 6 modules
-
Covers hidden instructions delivered through web pages, emails, documents, RAG sources and MCP tool outputs, data exfiltration chains, and architectural defences such as spotlighting, the dual-LLM pattern and CaMeL.
- Live Course
- 5 modules
Get in touch
