All templates By NappAI team CV Scorer
Flow in which a job position is specified and available CVs are scored based on each one's value for that job.
graph TD
%%{init: {'theme': 'mc','layout': 'elk'}}%%
LanggraphReactAgent-vy5ag[Agent]
style LanggraphReactAgent-vy5ag stroke:#a170ff
ChatOutput-h6ga5[<div><img src="/_astro/messages-square.BaSDmT6g.svg" style="height: 20px !important;width: 20px !important"/></div>Chat Output]
style ChatOutput-h6ga5 stroke:#a170ff
TextInput-wm8fq[<div><img src="/_astro/type.Dy26vmDy.svg" style="height: 20px !important;width: 20px !important"/></div>Job Description]
style TextInput-wm8fq stroke:#a170ff
GDriveFilesComponent-ozzfk[<div><img src="/_astro/google_drive.wKmDsV2c.svg" style="height: 20px !important;width: 20px !important"/></div>Google Drive File Manager]
style GDriveFilesComponent-ozzfk stroke:#a170ff
OpenAIModel-sjaw0[<div><img src="/_astro/openAI.BhmuxEs3.svg" style="height: 20px !important;width: 20px !important"/></div>OpenAI]
style OpenAIModel-sjaw0 stroke:#a170ff
GDriveFilesComponent-ozzfk -.- LanggraphReactAgent-vy5ag
linkStyle 0 stroke:#a170ff
OpenAIModel-sjaw0 -.- LanggraphReactAgent-vy5ag
linkStyle 1 stroke:#a170ff
LanggraphReactAgent-vy5ag -.- ChatOutput-h6ga5
linkStyle 2 stroke:#a170ff
TextInput-wm8fq -.- LanggraphReactAgent-vy5ag
linkStyle 3 stroke:#a170ff
CV Scorer Workflow Documentation
🧩 Overview
The CV Scorer workflow automatically evaluates candidates’ resumes against a specified job description. It gathers CVs from a Google Drive folder, extracts key information, scores each candidate based on defined criteria, and delivers a ranked summary in a chat‑style interface. The process streamlines talent screening, saving recruiters time and ensuring consistent, objective assessment.
⚙️ Main Features
- Job description ingestion – accepts a full‑text description of the position.
- Google Drive integration – retrieves all CV files from a selected folder.
- AI‑powered analysis – employs a large language model to parse CVs and calculate compatibility scores.
- Rule‑based scoring – applies technical, experience, education, and soft‑skill criteria.
- Chat‑style output – presents results in a readable, structured message.
- Extensible tool set – allows additional tools to be attached to the agent for future expansion.
🔄 Workflow Steps
| Component Name | Role in the Workflow | Key Inputs | Key Outputs |
|---|---|---|---|
| Job Description Input | Captures the textual description of the role to be evaluated. | Raw job description text. | Formatted job description message. |
| Recruitment Agent | Orchestrates the scoring process, coordinating the language model and tools. | Job description message, list of available tools, language‑model reference. | Final ranked summary message. |
| Google Drive File Manager | Retrieves CV files from a specified Drive folder for processing. | Folder identifier (e.g., “CVs”) and optional file‑type filter. | List of file contents or base‑64‑encoded data for each CV. |
| OpenAI Model | Generates natural‑language outputs and JSON‑structured data for each CV. | CV content or extracted text, scoring prompt. | JSON object containing extracted fields and a numeric compatibility score. |
| Chat Output | Displays the agent’s final response in the playground chat interface. | Ranked summary message. | Human‑readable chat message. |
🧠Notes
- Tool integration – The Recruitment Agent automatically registers the Drive File Manager as a tool; no manual wiring is required.
- API credentials – Secure keys for OpenAI and Google Drive must be configured in the environment; the workflow does not expose them.
- Execution limits – The agent is capped at 10 iterations and a 10‑second runtime per cycle to prevent runaway processes.
- Fallback model – A secondary language model can be enabled; if the primary model fails, the fallback will take over without interrupting the workflow.
- Data privacy – All CVs are processed in memory and are not stored persistently unless explicitly configured.
- JSON output – When
json_modeis enabled, the OpenAI Model will return strictly JSON, which the agent uses for scoring logic. - Scalability – The Drive File Manager supports batch processing; for very large folders, consider pagination or selective retrieval.