At Independent / Remote, I design and audit multi-turn conversational scenarios to evaluate LLM reasoning, contextual continuity, and instruction-following fidelity.
I construct and validate JSON payloads for function-calling tasks, checking parameter extraction across Calendar, Maps, and Search integrations. I also analyze responses for hallucinations, redundant tool calls, and logical failures.
I developed an Agentic Function-Calling Benchmark Suite with 30+ scenarios covering ambiguous requests, missing parameters, and mock API failures. My Prompt Optimization Matrix reduced JSON formatting and schema validation errors in model outputs by over 25%.
At Cisco Networking Academy, I configured and failure-tested network topologies, troubleshot connectivity and hardware faults, and documented setups and procedures. I completed CCNA coursework there.
