Skip to content

jackyilmaz/qa-orchestrator-platform

BranchesTags

Repository files navigation

QA Orchestrator Platform

Java Spring Boot API Contract LLM DB Copilot

QA Orchestrator Platform is an AI-powered QA decision engine covering the full QA lifecycle — from ticket creation to release — integrated with Microsoft Copilot Studio and Power Automate as a 7-agent system.

Live

URL
Dashboard https://qa-orchestrator-service.onrender.com
API https://qa-orchestrator-service.onrender.com/qa/api/v1/qa/analyze
Health https://qa-orchestrator-service.onrender.com/qa/health

Full QA Lifecycle

Ticket → "In Progress" → QA pipeline runs automatically
                       → Requirement Analysis
                       → Test Design
                       → Automation Decision
                       → Risk Analysis
                       → Bug Report Template
                       → Saved to DB + Jira comment

Ticket → "Done"       → QA Release Summary generated
                       → Verdict: APPROVED / APPROVED WITH RISK / RELEASED WITHOUT FULL COVERAGE
                       → Saved to DB + Jira comment

System Architecture

User / Copilot Studio / Power Automate / Jira Webhook
            │
            ▼
    QA Orchestrator API
            │
            ▼
     Spring Boot Service
       │           │           │
       ▼           ▼           ▼
  Jira REST   Azure OpenAI  Neon PostgreSQL

Multi-Agent Copilot Studio

7 dedicated agents — each returns only its relevant stage output. No full pipeline dumps in chat.

Agent Trigger Example Dedicated Endpoint
agent_requirement_analyzer "What are the requirements for project-5" POST /qa/api/v1/agent/requirements
agent_test_case_generator "Create test cases for project-6" POST /qa/api/v1/agent/testcases
agent_risk_predictor "What is the risk for project-7" POST /qa/api/v1/agent/risk
agent_automation_builder "Automation strategy for project-8" POST /qa/api/v1/agent/automation
agent_bug_reporter "Create a bug report for project-5" POST /qa/api/v1/agent/bugreport
agent_intelligence_summary "Give me a QA summary" GET /qa/api/v1/intelligence/summary
agent_release_summary "Which tickets have been released" GET /qa/api/v1/intelligence/released/summary

Issue key normalization is automatic — "project-8" → "PROJ-8".

LLM Provider Support

Provider Env Var Model Notes
Azure OpenAI (active) LLM_PROVIDER=azure GPT-4o Current active provider
Groq LLM_PROVIDER=groq Llama 3.3 70B Fast, free tier alternative
AWS Bedrock LLM_PROVIDER=aws Claude 3.5 Sonnet Enterprise, AWS ecosystem

Switch providers via LLM_PROVIDER env var — no code changes required.

QA Analysis Pipeline

Stage Output
Requirement Analysis clarifiedRequirements, edgeCases, openQuestions, scope
Test Design testScenarios, testCases (UI/API/E2E)
Automation Decision automationRecommendation, coverageSplit, framework
Risk Analysis riskScore (0–100), riskLevel, releaseRecommendation
Bug Report title, severity, reproductionSteps, impactSummary
Release Summary APPROVED / APPROVED WITH RISK / RELEASED WITHOUT FULL COVERAGE

Intelligence Dashboard

https://qa-orchestrator-service.onrender.com

  • Metrics: total analyses, avg risk, blocked, released
  • Risk distribution and release decision charts
  • Recent analyses — searchable, filterable by risk and release decision
  • Blocked tickets — searchable
  • Released tickets with QA verdicts — searchable

Jira Webhook

Status Action
In Progress Full QA analysis pipeline
Done QA release summary generation

Setup: Jira → System → WebHooks → URL: https://qa-orchestrator-service.onrender.com/qa/webhook/jira → Event: Issue updated → JQL: project = PROJ

API Endpoints

Core

Method Path Description
GET / Redirects to dashboard
GET /qa/health Health + intelligence summary
GET /qa/dashboard Intelligence dashboard
POST /qa/api/v1/qa/analyze Full pipeline analysis
POST /qa/webhook/jira Jira webhook receiver

Agent Endpoints (Copilot Studio)

Method Path Description
POST /qa/api/v1/agent/requirements Requirements stage only
POST /qa/api/v1/agent/testcases Test cases stage only
POST /qa/api/v1/agent/risk Risk analysis stage only
POST /qa/api/v1/agent/automation Automation strategy only
POST /qa/api/v1/agent/bugreport Bug report only

History & Intelligence

Method Path Description
GET /qa/api/v1/history Last 10 analyses
GET /qa/api/v1/history/{issueKey} Per-issue history
GET /qa/api/v1/intelligence/summary Aggregated intelligence summary
GET /qa/api/v1/intelligence/high-risk HIGH risk analyses
GET /qa/api/v1/intelligence/blocked Blocked analyses
GET /qa/api/v1/intelligence/released Released tickets with QA verdicts
GET /qa/api/v1/intelligence/released/summary Released summary (Copilot-friendly)
GET /qa/api/v1/intelligence/trends Risk trends across re-analyzed issues
GET /qa/api/v1/intelligence/trends/{issueKey} Risk timeline for specific issue
GET /qa/api/v1/intelligence/reanalyzed Most re-analyzed issues

Environment Variables

Variable Required Description
JIRA_BASE_URL Yes Jira instance URL
JIRA_EMAIL Yes Jira account email
JIRA_API_TOKEN Yes Jira API token
AZURE_OPENAI_KEY Yes (Azure) Azure OpenAI key
AZURE_OPENAI_ENDPOINT Yes (Azure) Azure OpenAI endpoint
AZURE_OPENAI_DEPLOYMENT No Model deployment name (default: gpt-4o)
GROQ_API_KEY Yes (Groq) Groq API key
AWS_ACCESS_KEY Yes (AWS) AWS access key
AWS_SECRET_KEY Yes (AWS) AWS secret key
AWS_REGION No AWS region (default: us-east-1)
LLM_PROVIDER No azure / groq / aws (default: azure)
SPRING_DATASOURCE_URL Yes Neon PostgreSQL JDBC URL
SPRING_DATASOURCE_USERNAME Yes PostgreSQL username
SPRING_DATASOURCE_PASSWORD Yes PostgreSQL password
JIRA_COMMENT_ENABLED No Post analysis to Jira comment (default: false)

Local Development

export JIRA_BASE_URL=https://your-domain.atlassian.net
export [email protected]
export JIRA_API_TOKEN=your-jira-api-token
export AZURE_OPENAI_KEY=your-azure-openai-key
export AZURE_OPENAI_ENDPOINT=https://your-resource.openai.azure.com/
export AZURE_OPENAI_DEPLOYMENT=gpt-4o
export LLM_PROVIDER=azure

./mvnw spring-boot:run
curl -X POST http://localhost:10000/qa/api/v1/qa/analyze \
-H "Content-Type: application/json" \
-d '{"issueKey":"PROJ-4"}'

Tech Stack

Layer Technology
Backend Java 17, Spring Boot 3, Maven
LLM Azure OpenAI (GPT-4o) — active; Groq and AWS Bedrock supported
Database Neon PostgreSQL (free tier, no expiry, AWS US East 1)
Infrastructure Docker, Render Cloud
Integrations Jira REST API, Microsoft Copilot Studio, Power Automate

Roadmap

Phase Status Description
1 ✅ Foundation — backend, Jira, pipeline, deployment
2 ✅ LLM Intelligence — all 5 stages AI-powered
3 ✅ Observability — logging, health, error handling
4 ✅ Hardening — validation, timeouts, security
5 ✅ Intelligence layer — PostgreSQL, dashboard, history
6 ✅ Full lifecycle — webhook, release summary, LLM providers
7 ✅ Copilot Studio v2 — intelligence + release topics
8 ✅ Dashboard — search, filtering, record counts
9 ✅ Multi-tenant foundation — TenantConfig, architecture ready
10 ✅ Risk trend analysis — timeline, trends, reanalyzed issues
11 ✅ Multi-Agent Copilot Studio — 7 dedicated agents, focused endpoints
12 ✅ Database migration — Render → Neon PostgreSQL
13 📋 Production hardening — Azure/AWS, SOC2, rate limiting

About

AI-powered QA analysis platform that converts Jira issues into structured testing strategies, automation guidance, and release risk analysis.

qa-orchestrator-service.onrender.com

Topics

docker qa jira ai spring-boot render test-automation automation-framework qa-platform sdet power-automate ai-testing copilot-studio

Resources

Readme

License

View license
Activity

Stars

1 star

Watchers

0 watching

Forks

0 forks
Report repository

Releases

No releases published

Packages

 
 
 

Contributors

Languages