Building an AI agent that chases supplier quotes: what we learned about RFQ emails

说明与用途

AI 代理 Dov 自动化发送结构化 RFQ、跟进、从多种格式提取并规范报价数据，并进行到岸成本比较，最终由人工批准。

问题

Automating the procurement process of sending RFQs, chasing supplier replies, and parsing messy quotes into comparable data while avoiding missing or incorrect information.

结果

The AI agent reduced manual effort, ensured consistent follow‑ups, produced structured comparison tables, and maintained a human approval step, improving procurement efficiency and data quality.

工作流

• 1. Draft a structured RFQ with part description, material, tolerances, finish, quantity tiers, incoterms, destination, lead time, and a numbered list of questions.
• 2. Send the RFQ via email.
• 3. Wait 2 business days, then send a first follow‑up; wait 4 more days for a second follow‑up; send a final nudge asking to close the request.
• 4. Receive supplier replies in PDF, Excel, text, or photo format.
• 5. Extract text and tables from the received format.
• 6. Map extracted data onto a fixed schema (unit price, currency, MOQ, tooling cost, lead time, payment terms, validity date).
• 7. Flag any missing fields and send targeted follow‑ups to suppliers.
• 8. Normalize each quote to an estimated landed cost per unit, including freight, duties, tooling amortization, payment terms, and lead‑time cost.
• 9. Present a comparison table to a human reviewer for approval before any purchase is made.

证据

Suppliers answer the question you ask. If your RFQ says "need 5,000 aluminium brackets, please quote", you will get five different interpretations.

来源

Most quotes arrive after a nudge, not after the first email.

来源

Quotes arrive as PDFs, Excel files, inline text, and the occasional photographed price list.

来源

The rule: never let the model silently fill a gap. A blank cell is information. An invented price is a lawsuit.

来源

Agents are great at finding suppliers, sending clean RFQs, chasing, and building the comparison table. They should not unilaterally commit money.

来源