How to Build a Resume‑Screening AI Agent Workflow with FastGPT

FastGPT is presented as a convenient platform for constructing AI applications. The author uses it to create a human‑resources assistant that analyses resume content with natural‑language processing, extracts key fields such as education, work experience, and skill match, and generates an objective score report.

The traditional resume‑screening process is described as time‑consuming, prone to subjective bias, and error‑prone when reviewers become fatigued. The AI‑driven solution can analyze hundreds of resumes within seconds, reducing screening time by more than 70 % and applying a uniform scoring rubric to eliminate bias.

Step 1 – Create a Feishu application : Access the Feishu Open Platform (https://open.feishu.cn), register, and obtain the App ID and App Secret.

Step 2 – Enable required permissions : In the permission management section, enable the Cloud Docs and Multi‑dimensional Table identities for the app. Step 3 – Configure the FastGPT workflow : The author uploads a JSON workflow configuration (available on request) into FastGPT, which defines the sequence of AI calls, data extraction, and table updates. A screenshot of the workflow editor is shown. Step 4 – Set environment variables : Before running the workflow, the required variables (e.g., FastGPT API key, Feishu credentials) are entered. Step 5 – Upload resumes and execute : Users upload PDF or Word resumes, trigger the workflow, and the system extracts information, applies the scoring prompts, and writes results to a Feishu multi‑dimensional table. Step 6 – Review results : The populated table displays scores for personal ability, job‑match percentage, education, and other dimensions, along with execution logs. The core of the workflow (module 2) calls multiple AI models, aggregates their outputs, and formats the final report. Prompt examples used for scoring are provided verbatim:

# 角色
您具备精准匹配简历与岗位需求的能力，尤其擅长对候选人的个人综合能力进行专业评估与量化评分
## 评分标准
### 1. 逻辑与表达（满分10分）
简历结构：板块清晰、重点突出（5分），混乱扣3分
语言表述：无错别字/语法错误（3分），每处错误扣0.5分
量化描述：80%以上内容含数据支撑（+2分）
## 2. 综合素质（满分15分）
团队协作：描述团队项目中的角色贡献（5分）
问题解决：案例体现分析能力（如“独立解决技术故障”，5分）
责任心/抗压能力：通过实习/项目经历佐证（5分）
# 限制
仅输出数字，不要输入任何无关内容
满分为25分，请不要超过25分

For job‑match evaluation, a separate prompt defines a weighted scoring model (40 % skill match, 30 % experience relevance, 20 % soft‑skill fit, 10 % potential). The AI returns a percentage score. By automating resume parsing, scoring, and data entry, the workflow dramatically speeds up the initial screening stage, allowing HR professionals to focus on higher‑value interview and decision‑making tasks.