Scope: a multi-step LLM pipeline or agent wired into your product: prompt chains, embeddings + clustering, ranking, LLM-as-judge evaluation, tool-calling agents (LangGraph), MCP servers, and queue-backed workers with retries (BullMQ / Python).
Deliverables: a working pipeline in your repo, an eval set with before/after accuracy numbers, retry/failure handling and logs, a cost-per-run estimate, and a short README.
Starting price: $2,200 for a 1-week scoped pipeline (or $55/hr).
Scope: a multi-step LLM pipeline or agent wired into your product: prompt chains, embeddings + clustering, ranking, LLM-as-judge evaluation, tool-calling agents (LangGraph), MCP servers, and queue-backed workers with retries (BullMQ / Python).
Deliverables: a working pipeline in your repo, an eval set with before/after accuracy numbers, retry/failure handling and logs, a cost-per-run estimate, and a short README.
Starting price: $2,200 for a 1-week scoped pipeline (or $55/hr).