AI Prompt Engineer & Developer (Freelance) — Self-Employed / Various Enterprise Clients (Remote)
Freelance AI Prompt Engineer and Developer evaluating LLM outputs for strict user personas and improving constraint adherence to reduce hallucinations. Work included testing responses for structured JSON extraction with zero errors and benchmarking model extraction of key fields from incoming emails. Responsibilities also covered prompt refinements for persona/tone guardrails and quality checks before automated actions or routing decisions. • Evaluated public health agent outputs against personas and formatting constraints • QA retail Telegram support agent outputs by validating retrieval-augmented responses • Curated email test suites to benchmark structured JSON extraction accuracy • Refined few-shot/system prompts and edge-case testing for conversational agents