Skip to main content

​
Endpoint

GET /api/interview/sessions/{sessionId}/report
Generates a comprehensive interview evaluation report using AI analysis. The report includes overall scoring, category-based breakdowns, detailed feedback for each answer, strengths identification, improvement suggestions, and reference answers for comparison.
This endpoint should only be called after the interview session is completed (status: COMPLETED). Calling it on an incomplete session may result in partial or invalid results.

​
Path Parameters

​
sessionId
string
required
The unique session identifier for the completed interview session.

​
Response

​
code
integer
Response status code. 200 indicates success.
​
message
string
Response message. Returns "success" when report is generated.
​
data
object
The comprehensive interview evaluation report.

​
Example Request

curl https://api.example.com/api/interview/sessions/a1b2c3d4-e5f6-4789-g0h1-i2j3k4l5m6n7/report

​
Example Response

{
  "code": 200,
  "message": "success",
  "data": {
    "sessionId": "a1b2c3d4-e5f6-4789-g0h1-i2j3k4l5m6n7",
    "totalQuestions": 8,
    "overallScore": 82,
    "categoryScores": [
      {
        "category": "项目经历",
        "score": 85,
        "questionCount": 2
      },
      {
        "category": "Spring",
        "score": 88,
        "questionCount": 3
      },
      {
        "category": "MySQL",
        "score": 75,
        "questionCount": 2
      },
      {
        "category": "Redis",
        "score": 78,
        "questionCount": 1
      }
    ],
    "questionDetails": [
      {
        "questionIndex": 0,
        "question": "请介绍一下你在电商平台项目中如何使用Spring Cloud构建微服务架构的？",
        "category": "项目经历",
        "userAnswer": "在电商平台项目中，我们使用Spring Cloud构建了微服务架构。主要包括以下组件：使用Eureka作为服务注册中心，Zuul作为API网关，Feign实现服务间调用，Hystrix提供熔断降级功能。我们将系统拆分为订单服务、商品服务、用户服务等多个微服务，通过配置中心统一管理配置，使用消息队列实现异步解耦。",
        "score": 85,
        "feedback": "回答全面，涵盖了Spring Cloud的核心组件。提到了服务注册、API网关、服务调用、熔断降级等关键点。服务拆分思路清晰。建议补充具体的技术挑战和解决方案，比如如何处理分布式事务、如何保证服务间调用的幂等性等。"
      },
      {
        "questionIndex": 1,
        "question": "你提到使用Hystrix实现熔断降级，能具体说说在什么场景下会触发熔断吗？你们是如何配置熔断参数的？",
        "category": "Spring",
        "userAnswer": "当下游服务响应时间过长或者频繁失败时会触发熔断。我们设置了超时时间为2秒，失败率超过50%时触发熔断。熔断后会返回默认值或者走降级逻辑。",
        "score": 78,
        "feedback": "理解了熔断的基本概念和触发条件。提到了超时时间和失败率阈值。但是对熔断的具体机制理解不够深入，建议补充：1) 熔断器的三种状态（关闭、打开、半开）及其转换条件；2) 时间窗口的概念；3) 最小请求数阈值；4) 半开状态下的恢复机制。"
      }
    ],
    "overallFeedback": "候选人展现了扎实的Java后端开发经验，对Spring Cloud微服务架构有较好的理解和实践经验。在项目经历和Spring框架方面表现优秀，能够清晰描述技术方案。在数据库优化和缓存设计方面有一定经验，但深度还可以提升。建议加强对底层原理的理解，特别是分布式系统的理论基础、MySQL的执行计划分析、Redis的持久化和高可用方案。整体来说，适合中高级Java开发岗位。",
    "strengths": [
      "深入理解Spring Cloud微服务架构，有实际项目经验",
      "熟悉分布式系统设计，了解服务治理的各个方面",
      "有大规模项目的实战经验，处理过高并发场景",
      "技术表达清晰，能够系统性地阐述技术方案"
    ],
    "improvements": [
      "需要加强对熔断降级机制的深入理解，包括状态转换和参数调优",
      "建议深入学习MySQL的执行计划和索引优化策略",
      "Redis的持久化机制和高可用方案需要进一步掌握",
      "可以补充更多关于分布式事务的实践经验"
    ],
    "referenceAnswers": [
      {
        "questionIndex": 0,
        "question": "请介绍一下你在电商平台项目中如何使用Spring Cloud构建微服务架构的？",
        "referenceAnswer": "在电商平台项目中，我们采用Spring Cloud全家桶构建了微服务架构。具体技术选型和实施方案如下：\n\n1. 服务注册与发现：使用Eureka Server作为注册中心，各微服务启动时自动注册，实现服务的动态发现和负载均衡。\n\n2. API网关：使用Zuul/Gateway统一对外提供服务，实现路由转发、权限校验、限流熔断等功能。\n\n3. 服务间调用：使用Feign声明式服务调用，集成Ribbon实现客户端负载均衡。\n\n4. 熔断降级：使用Hystrix实现服务保护，当依赖服务异常时自动降级，保证系统整体可用性。\n\n5. 配置中心：使用Spring Cloud Config统一管理各环境配置，支持动态刷新。\n\n6. 链路追踪：集成Sleuth+Zipkin实现分布式链路追踪，便于问题定位。\n\n7. 服务拆分：按业务领域拆分为订单服务、商品服务、用户服务、支付服务等，每个服务独立部署、独立扩展。\n\n在实施过程中，我们还解决了分布式事务（Seata）、服务间认证（JWT）、日志收集（ELK）等关键问题，整体架构支撑了日均10万+订单的业务量。",
        "keyPoints": [
          "服务注册与发现机制（Eureka）",
          "API网关功能（路由、鉴权、限流）",
          "服务间调用和负载均衡（Feign + Ribbon）",
          "熔断降级方案（Hystrix）",
          "配置中心应用（Config）",
          "链路追踪实现（Sleuth + Zipkin）",
          "服务拆分原则和实践",
          "分布式问题解决方案（事务、认证等）"
        ]
      },
      {
        "questionIndex": 1,
        "question": "你提到使用Hystrix实现熔断降级，能具体说说在什么场景下会触发熔断吗？你们是如何配置熔断参数的？",
        "referenceAnswer": "Hystrix熔断机制的工作原理和配置如下：\n\n熔断触发场景：\n1. 服务超时：依赖服务响应时间超过设定阈值\n2. 服务异常：依赖服务抛出异常\n3. 线程池/信号量满：资源耗尽无法处理请求\n\n熔断器状态：\n1. CLOSED（关闭）：服务正常，请求正常通过\n2. OPEN（打开）：达到熔断条件，快速失败，执行降级逻辑\n3. HALF_OPEN（半开）：尝试恢复，允许部分请求通过测试\n\n关键配置参数：\n1. circuitBreaker.requestVolumeThreshold: 20  // 时间窗口内最小请求数\n2. circuitBreaker.errorThresholdPercentage: 50  // 错误率阈值50%\n3. circuitBreaker.sleepWindowInMilliseconds: 5000  // 熔断后5秒进入半开状态\n4. execution.isolation.thread.timeoutInMilliseconds: 2000  // 超时时间2秒\n\n降级策略：\n1. 返回默认值或缓存数据\n2. 返回友好提示信息\n3. 调用备用服务\n\n在实际项目中，我们根据不同服务的重要性和性能特征，设置了差异化的熔断参数，并通过监控面板实时观察熔断情况。",
        "keyPoints": [
          "熔断触发条件（超时、异常、资源耗尽）",
          "熔断器三种状态及转换逻辑",
          "时间窗口和最小请求数概念",
          "错误率阈值配置",
          "半开状态的恢复机制",
          "降级策略的具体实现",
          "差异化配置和监控"
        ]
      }
    ]
  }
}

​
Report Structure

The interview report provides multi-level insights:
1

Overall Score

High-level assessment (0-100) indicating the candidate’s overall performance.
2

Category Breakdown

Scores grouped by technical category (Java, Spring, databases, etc.) to identify strengths and weaknesses.
3

Question-level Details

Individual feedback for each answer explaining what was good and what could be improved.
4

Comprehensive Feedback

Overall assessment summarizing performance, technical depth, and readiness for the role.
5

Actionable Insights

Specific strengths to leverage and improvements to focus on for growth.
6

Reference Materials

Model answers and key points to help candidates learn and improve.

Export Report

Download the report as a formatted PDF

Get Details

Retrieve complete interview session details

Delete Session

Remove the interview session and report

Create Session

Start a new interview session

​
Additional Operations

​
Get Interview Details

Retrieve the complete interview session including questions, answers, and evaluation details. 
GET /api/interview/sessions/{sessionId}/details
​
sessionId
string
required
The session identifier.
Example: 
curl https://api.example.com/api/interview/sessions/a1b2c3d4-e5f6-4789-g0h1-i2j3k4l5m6n7/details
Response: Returns InterviewDetailDTO with complete session history, all questions, answers, and evaluation data.

​
Error Responses

​
code
integer
Error code. Non-200 values indicate an error.
​
message
string
Error message describing what went wrong.
​
data
null
Always null for error responses.

​
Common Errors

CodeMessageDescription
404Session not foundInvalid session ID
400Interview not completedCannot generate report for incomplete interview
500AI service errorFailed to generate evaluation
500Server errorInternal server error

​
Best Practices

Report generation may take a few seconds. Cache the report once generated to avoid regenerating:
let cachedReport = null;

async function getReport(sessionId) {
  if (cachedReport && cachedReport.sessionId === sessionId) {
    return cachedReport;
  }
  
  const response = await fetch(`/api/interview/sessions/${sessionId}/report`);
  cachedReport = response.data;
  return cachedReport;
}
Show loading state while the AI generates the report:
showLoadingMessage('Analyzing your answers...');
const report = await generateReport(sessionId);
hideLoadingMessage();
displayReport(report);
Use charts to make category scores more digestible:
// Radar chart for category comparison
const categories = report.categoryScores.map(c => c.category);
const scores = report.categoryScores.map(c => c.score);
renderRadarChart(categories, scores);
Emphasize reference answers and key points for educational value:
report.referenceAnswers.forEach(ref => {
  displayReferenceAnswer(ref.question, ref.referenceAnswer);
  displayKeyPoints(ref.keyPoints);
});
Provide export options immediately after showing the report:
displayReport(report);
showExportButton(`/api/interview/sessions/${sessionId}/export`);

​
Report Interpretation

Score Ranges:
  • 90-100: Excellent performance - Strong candidate with deep knowledge
  • 75-89: Good performance - Solid understanding with room to grow
  • 60-74: Adequate performance - Basic knowledge, needs improvement
  • Below 60: Needs significant improvement - Consider additional training
Category Scores help identify specific areas of strength and weakness for targeted improvement.

​
AI Evaluation Features

The report generation uses advanced AI to:
  • Evaluate technical accuracy and depth of answers
  • Assess communication clarity and structure
  • Identify specific knowledge gaps
  • Generate personalized improvement suggestions
  • Create relevant reference answers
  • Provide actionable learning paths
The report is designed to be educational. Encourage candidates to review reference answers and key points to learn from the interview experience.

​
Session State Update

After generating a report, the session status automatically changes from COMPLETED to EVALUATED. This prevents duplicate report generation and marks the interview as fully processed.

Build docs developers (and LLMs) love

Get started for freeTalk to us