🚀 Building OSSI — An AI-Powered Open Source Intelligence System with Kestra

Transforming GitHub issues into contributor intelligence using AI + Workflow Orchestration

Open source projects are growing faster than ever.

Every day repositories receive:

• Bug reports
• Feature requests
• Contributor discussions
• Engineering questions
• Documentation improvements
• Infrastructure problems

But there’s a major challenge:

Maintainers are overwhelmed.

Important issues get buried.
Contributors struggle to find meaningful tasks.
Stale issues pile up.
Project health slowly declines.

So I built:

OSSI — Open Source Signal Intelligence System

An AI-powered orchestration platform built using Kestra that transforms GitHub repositories into actionable contributor intelligence.

🔗 GitHub Repository:
https://github.com/Mohit5Upadhyay/ossi-intel-orchestrator

🧠 What is OSSI?

OSSI stands for:

Open Source Signal Intelligence System

It’s an autonomous workflow orchestration platform that:

✅ Monitors GitHub repositories
✅ Fetches live GitHub issues
✅ Detects stale issues
✅ Uses AI to analyze engineering complexity
✅ Prioritizes issues automatically
✅ Recommends contributor actions
✅ Generates intelligence reports
✅ Sends automated engineering summaries via email
✅ Runs continuously on schedules using Kestra

Instead of manually reading hundreds of issues, OSSI creates an intelligent engineering layer over open-source repositories.

🤔 Why Does OSSI Matter?

Open source maintainers deal with a serious scaling problem.

As repositories grow:

• Issue backlogs explode
• Contributors become confused
• Duplicate issues increase
• Stale tickets accumulate
• Prioritization becomes difficult

And contributors face problems too:

Contributors struggle with:

• Finding beginner-friendly issues
• Understanding issue complexity
• Knowing project priorities
• Discovering impactful tasks
• Understanding technical context

OSSI solves this by turning repositories into structured engineering intelligence systems.

⚡ What OSSI Actually Does

The workflow continuously scans repositories and transforms raw GitHub data into actionable insights.

Example Intelligence Generated

OSSI automatically identifies:

• High-priority issues
• Beginner-friendly tasks
• Advanced engineering problems
• Potential stale issues
• Contributor recommendations
• Root cause analysis
• Suggested implementation steps

Example AI-generated output:

Priority: High

Difficulty: Intermediate

Good First Issue: Yes

Root Cause:
Missing validation layer causing inconsistent API responses.

Quick Fix Approach:
1. Add schema validation
2. Implement middleware checks
3. Add automated tests

This turns raw GitHub issues into contributor-ready engineering tasks.

🚀 Why Kestra Was the Perfect Choice

This entire system is powered by Kestra.

And honestly — Kestra completely changed how I think about automation.

Most automation tools feel like:

• Task runners
• Cron jobs
• Simple scripting systems

But Kestra feels like:

• Infrastructure orchestration
• Workflow operating systems
• AI pipeline orchestration
• Distributed automation architecture

🔥 What Makes Kestra Powerful

1️⃣ Everything is Declarative

The entire orchestration pipeline is written in YAML.

2️⃣ Built-in Scheduling

OSSI runs every 6 hours automatically.

triggers:
  - id: scheduled_ossi_scan
    type: io.kestra.plugin.core.trigger.Schedule

    cron: "0 */6 * * *"

    timezone: "Asia/Kolkata"

No external schedulers needed.

3️⃣ Parallel Processing

OSSI processes repositories dynamically using ForEach.

- id: process_repositories
  type: io.kestra.plugin.core.flow.ForEach

  values: "{{ inputs.repositories }}"

This allows multi-repository intelligence generation.

4️⃣ Native API Integrations

Kestra makes API orchestration incredibly easy.

Example GitHub issue fetching:

- id: fetch_open_issues
  type: io.kestra.plugin.core.http.Request

  method: GET

  uri: "{{ vars.github_api }}?q=repo:{{ taskrun.value }}+is:issue+is:open"

  headers:
    Authorization: "Bearer {{ inputs.github_pat }}"

This is extremely clean compared to building everything manually.

5️⃣ AI Workflow Orchestration

One of the most powerful parts:

Kestra orchestrates AI systems beautifully.

OSSI sends repository issue data into AI models for:

• Engineering analysis
• Contributor recommendations
• Priority ranking
• Root cause reasoning
• Issue classification

This is where orchestration becomes much more than automation.

🏗️ OSSI Workflow Architecture

Here’s the full orchestration pipeline:

graph TD
    A[Schedule Trigger] --> B[Process Repositories]
    B --> C[Fetch GitHub Issues]
    C --> D[Process Issue Data]
    D --> E[AI Contributor Analysis]
    E --> F[Generate Intelligence Report]
    F --> G[Send Email Report]
    G --> H[Workflow Completed]

🔍 Deep Dive Into the Workflow

1️⃣ Workflow Trigger

The workflow starts automatically every 6 hours.

triggers:
  - id: scheduled_ossi_scan
    type: io.kestra.plugin.core.trigger.Schedule

    cron: "0 */6 * * *"

    timezone: "Asia/Kolkata"

This transforms OSSI into a continuously running intelligence system.

2️⃣ Processing Multiple Repositories

OSSI supports multiple repositories dynamically.

inputs:

  - id: repositories
    type: ARRAY
    itemType: STRING

Example repositories:

defaults:
  - "kestra-io/kestra"
  - "open-metadata/OpenMetadata"

3️⃣ GitHub Issue Intelligence

The workflow fetches live issues directly from GitHub APIs.

- id: fetch_open_issues
  type: io.kestra.plugin.core.http.Request

  method: GET

  uri: "{{ vars.github_api }}?q=repo:{{ taskrun.value }}+is:issue+is:open"

This creates a real-time engineering data stream.

4️⃣ Data Processing Using Shell + jq

After fetching issues, OSSI processes repository data.

commands:
  - |
    cat repo_issues.json | jq -r '
      .items[]
      | "
      Issue:
      #\(.number)

      Title:
      \(.title)
      "
    '

This stage transforms raw API responses into structured engineering summaries.

5️⃣ Stale Issue Detection

OSSI automatically identifies neglected issues.

select(.comments < 2)

This helps maintainers:

• Reduce backlog clutter
• Improve issue hygiene
• Re-engage contributors

Small automation.
Massive operational value.

6️⃣ AI Contributor Analysis

This is the brain of OSSI.

The workflow sends issue summaries into an AI model.

- id: ai_contributor_analysis
  type: io.kestra.plugin.core.http.Request

The AI then generates:

• Contributor recommendations
• Priority analysis
• Root cause insights
• Engineering reasoning
• Suggested implementation steps

This turns GitHub into an intelligent engineering platform.

7️⃣ 📧 SMTP Email Configuration

After generating intelligence reports, OSSI automatically delivers them using SMTP email orchestration.

Kestra makes email automation extremely clean.

Workflow email task:

- id: contributor_intelligence_email
  type: io.kestra.plugin.email.MailSend

  host: smtp.gmail.com

  port: 465

  username: "YOUR_USER_EMAIL_HERE"

  password: "YOUR_APP_PASSWORD_HERE"

  from: "YOUR_USER_EMAIL_HERE"

  to: "RECIPIENT_EMAIL_HERE"

  subject: "🚀 OSSI Intelligence Report"

  transportStrategy: SMTPS

The workflow automatically sends:

AI contributor insights
Engineering summaries
Issue prioritization
Repository intelligence
Stale issue detection reports

directly into your inbox.

8️⃣ 🤖 GitHub Models Configuration

OSSI uses GitHub Models to generate contributor intelligence automatically.

The workflow sends processed GitHub issue summaries into gpt-4o using Kestra's HTTP orchestration capabilities.

Actual workflow configuration:

- id: ai_contributor_analysis
  type: io.kestra.plugin.core.http.Request

  method: POST

  uri: "https://models.inference.ai.azure.com/chat/completions"

  headers:
    Content-Type: application/json
    Authorization: "Bearer {{ inputs.github_pat }}"

Model configuration:

{
  "model": "gpt-4o",
  "temperature": 0.2,
  "top_p": 1.0
}

The AI model performs:

Issue prioritization
Contributor recommendations
Root cause analysis
Engineering impact analysis
Difficulty classification
Beginner issue detection

This transforms OSSI into an autonomous engineering intelligence system instead of just a monitoring workflow.

9️⃣ Automated Email Intelligence Reports

🔐 Gmail SMTP Setup

To enable email delivery:

Enable 2-Factor Authentication on your Google account
Generate a Google App Password:
https://myaccount.google.com/apppasswords

Then replace:

username: "YOUR_USER_EMAIL_HERE"

password: "YOUR_APP_PASSWORD_HERE"

with your actual credentials.

This allows OSSI to autonomously deliver engineering intelligence reports after every workflow execution.

Finally, OSSI sends beautifully structured engineering reports directly to maintainers.

The report contains:

• Repository intelligence
• Contributor insights
• Engineering priorities
• Issue recommendations
• AI-generated analysis

All automatically orchestrated through Kestra.

🖥️ Setting Up OSSI Locally

Now let’s actually run it.

🐳 Step 1 — Run Kestra with Docker

Download the official Kestra Docker Compose file.

Linux/macOS

curl -o docker-compose.yml \
https://raw.githubusercontent.com/kestra-io/kestra/develop/docker-compose.yml

Windows
Invoke-WebRequest -Uri "https://raw.githubusercontent.com/kestra-io/kestra/develop/docker-compose.yml" -OutFile "docker-compose.yml"

Once the file is downloaded, start Kestra using:

docker compose up -d

🧩 Step 2 — Import the OSSI Workflow

Clone the repository:

git clone https://github.com/Mohit5Upadhyay/ossi-intel-orchestrator

Open Kestra UI.

Go to:

• Flows
• Create Flow
• Paste YAML workflow

Save the workflow.

Done.

🔐 Step 3 — Configure GitHub Token & Email Config

Generate GitHub PAT:
https://github.com/settings/tokens

Required permissions:

• repo

Then configure:

• github_pat
• recipient_email

inside workflow inputs.

▶️ Step 4 — Execute Workflow

Run the workflow manually OR wait for the schedule trigger.

Kestra will now:

• Fetch issues
• Analyze repositories
• Generate intelligence
• Send reports

autonomously.

📊 Kestra’s Visualization is Incredible

One thing I absolutely loved:

Workflow Topology View

Kestra visualizes:

• Task relationships
• Dependencies
• Execution structure
• Processing stages

This becomes extremely useful for complex orchestration systems.

⏱️ Live Execution Tracking

Kestra also provides:

• Gantt execution charts
• Runtime visibility
• Retry monitoring
• Failure tracking
• Execution logs

This makes debugging workflows much easier.

🧠 What I Learned Building OSSI

This project taught me something important:

AI becomes MUCH more powerful when combined with orchestration.

Without orchestration:

• AI is isolated

With orchestration:

• AI becomes infrastructure

That realization completely changed my engineering mindset.

🚀 Why Developers Should Learn Workflow Orchestration

If you're interested in:

• AI Engineering
• DevOps
• Automation
• ETL Systems
• AI Agents
• Distributed Systems
• Event-driven architectures
• Data pipelines

then orchestration is a critical skill.

And Kestra is one of the best tools I’ve used for learning it.

💡 Projects You Can Build Using Kestra

After building OSSI, I realized Kestra can orchestrate almost anything.

Some ideas:

• AI code review systems
• Autonomous CI/CD intelligence
• DevOps monitoring pipelines
• AI documentation generators
• Security analysis workflows
• Multi-agent AI systems

Once you start thinking in orchestration pipelines —
you begin engineering systems differently.

🔥 Why OSSI Matters for Open Source

OSSI is not just automation.

It’s:

• Contributor enablement
• Engineering intelligence
• Repository analytics
• AI-powered prioritization
• Open source acceleration

It helps:

• Maintainers scale better
• Contributors onboard faster
• Communities stay healthier
• Engineering efforts become focused

And I think systems like this will become increasingly important for the future of open source.

🛠️ Technologies Used

Core Stack

• Kestra
• GitHub API
• GitHub Models
• YAML
• Shell Scripting
• jq
• SMTP Automation

Concepts

• Workflow orchestration
• AI pipelines
• ETL processing
• Contributor intelligence
• Scheduled automation
• Engineering analytics

🔗 Project Links

GitHub Repository

https://github.com/Mohit5Upadhyay/ossi-intel-orchestrator

Kestra

https://kestra.io

Kestra GitHub

https://github.com/kestra-io/kestra

🎯 Final Thoughts

Before this project, I thought automation meant:
“running scripts automatically”

Now I think of orchestration as:
“building autonomous engineering systems”

OSSI started as a workflow experiment.

But it evolved into:

• AI infrastructure
• contributor intelligence
• engineering automation
• orchestration architecture

And honestly...

This feels like just the beginning of AI-powered workflow systems.

If you’re learning:

• AI Engineering
• Automation
• DevOps
• Workflow Systems
• Open Source Infrastructure

Build orchestration projects.

They teach you how real engineering systems operate.

And Kestra is an incredible place to start.

🚀