DEV Community: Prashant Lakhera

🔥 DevOps Open Agent now supports MCP🔥

Prashant Lakhera — Tue, 30 Jun 2026 16:01:20 +0000

We just added Model Context Protocol (MCP) integration to DevOps Open Agent, our open-source, self-hostable platform for AI-powered DevOps troubleshooting.

What that means in practice:

✅ Connect any MCP server — GitHub Copilot MCP, observability tools, runbooks, custom APIs

✅ Ask MCP — ask questions in plain English and let the platform call the right tools (e.g., list open PRs in my repo)

✅ Enrich AI investigations — Kubernetes, AWS, Cloud Cost, and PR Reviewer can include MCP tool context before diagnosis

✅ Your keys, your infra — per-user settings, optional Bearer auth, self-hosted

MCP isn’t a replacement for your agents; it extends them. Same UI, same investigation pipeline, now with external tools wired in.

Open source · Self-hostable · No vendor lock-in

🔗 GitHub: https://lnkd.in/giri6MF5

🎥 Demo: https://lnkd.in/eBWcPRtj

📝 Also, if you're interested in learning GenAI for DevOps, I have upcoming batches starting in July

https://ideaweaver.ai/

🚀 100 Days of GenAI is back, this time for DevOps Engineers!🚀

Prashant Lakhera — Tue, 30 Jun 2026 02:11:58 +0000

A few years ago, I started the 100 Days of DevOps series, and it helped thousands of engineers strengthen their DevOps skills.

Now it's time for the next chapter.

I'm excited to announce 100 Days of GenAI for DevOps Engineers, a completely FREE learning series designed to help DevOps professionals understand, build, and deploy AI-powered applications.

Over the next 100 days, we'll cover everything from AI fundamentals to production-ready AI systems, including:

✅ What each word in ChatGPT actually means

✅ AI Inference & GPUs

✅ AWS Bedrock

✅ AI with Kubernetes

✅ LangChain & LangGraph

✅ Ollama & vLLM

✅ RAG & MCP

✅ CrewAI & n8n Agents

✅ LLM Fine-Tuning

✅ Building a Small Language Model (SLM) from scratch

✅ LLM & Agent Evaluation

✅ AI for DevOps and much more...

🎁 Bonus: For the first time, the complete course will be available in both English and Hindi, making it accessible to an even wider community of developers and DevOps engineers.

Whether you're a DevOps Engineer, Cloud Engineer, Platform Engineer, SRE, or simply curious about Generative AI, this series is designed to take you from the fundamentals to building real-world AI solutions.

📅 The journey begins this Saturday(July 4), and I'll be spending one hour every day building and learning together with the community.

The course is completely FREE.

👇 The links to join the English and Hindi courses are available in the comments/description.

👇 The links to join the English and Hindi courses
📝 if you're interested in learning GenAI for DevOps, I have upcoming batches starting in July
https://ideaweaver.ai/
🔗 100 Days of GenAI for DevOps - English: www.ideaweaver.ai/purchase?product_id=6786786
🔗 100 Days of GenAI for DevOps - Hindi : www.ideaweaver.ai/purchase?product_id=6786785
🔗 Youtube Playlist: 100 Days of GenAI for DevOps - English : https://www.youtube.com/playlist?list=PLH7qLdABWUfU
🔗 Youtube Playlist: 100 Days of GenAI for DevOps - Hindi: https://www.youtube.com/playlist?list=PLbzaYqTigmW8
📖 Medium Series: https://devopslearning.medium.com/100-days-of-devops-81faf13bf772
💻 GitHub Repository: https://github.com/100daysofdevops/100daysofdevops

Who's joining me for the next 100 days? 🚀

✅ DevOps Open Agent: Enterprise-grade feedback(PagerDuty) shipped ✅

Prashant Lakhera — Mon, 29 Jun 2026 16:46:28 +0000

The feedback kept coming: Great tool, but where are the enterprise-grade integrations?

Fair point. DevOps teams don’t live in a single UI. When AI finds a root cause, someone needs to know, in the systems they already trust.

👻 I’m more than happy to announce that DevOps Open Agent now supports PagerDuty integration alongside Slack.

What this means in practice

When an investigation or PR review completes with AI diagnosis, recommendations can flow straight into your on-call workflow:

📌 PagerDuty — trigger incidents via Events API v2 (routing key, severity mapping, dedup keys)

📌 Slack — post root cause, suggested fix, and validation steps to your channel

📌 Per-user settings — routing key, cooldown period, and per-agent toggles (K8s, AWS, Cloud Cost, PR Reviewer)

📌 Alert fatigue protection — configurable cooldown so proactive monitoring doesn’t spam your team

Reactive troubleshooting when you need it. Proactive schedules when you want it. Enterprise-style alerting when it matters.

Open source. Self-hosted. Your infra, your LLM, your on-call stack.

🔗 Try it: https://github.com/ideaweaver-ai/devops-open-agent

🔗 Watch the demo: https://youtu.be/3VT8MeSLt5s

Would love your feedback: what integration should we add next? (Teams? Opsgenie? ServiceNow?)

📝 Also, if you're interested in learning GenAI for DevOps, I have upcoming batches starting in July

https://ideaweaver.ai/

✅ From reactive to proactive: DevOps Open Agent just leveled up ✅

Prashant Lakhera — Mon, 29 Jun 2026 00:54:55 +0000

For a while, DevOps Open Agent has been great at one thing: you hit a problem, you run an investigation, and AI helps you find the root cause, evidence, and suggested fixes across Kubernetes, AWS, cloud cost, and PR review.

That’s reactive DevOps powerful, but still you pulling the trigger every time.

I’m more than happy to announce that it’s now a proactive agent too.

What’s new

✔️ Proactive Kubernetes schedules

✔️ Set recurring investigations hourly, daily, weekly, or custom cron and let the same AI pipeline run in the background. Cluster health checks, namespace focus, and optional AI diagnosis on autopilot.

✔️ Slack delivery where your team already works

✔️ When an investigation completes, AI recommendations (root cause, fix, validation steps) can land in your Slack channel from Kubernetes, AWS, Cloud Cost, or PR Reviewer.

Alert fatigue protection

📌 Proactive doesn’t mean noisy. Slack alerts are capped at one per hour per user by default. Scheduled runs still complete and show up in the UI, you just don’t get spammed.

Plus, the platform keeps growing

➡️ AWS Lambda, EC2, VPC & S3 focused investigations

➡️ OpenRouter support (100+ models, one API key)

➡️ Self-hosted, open source, your LLM of choice

Reactive when you need it. Proactive when you want peace of mind.

🔗 Try it: https://lnkd.in/giri6MF5

🔗 Watch the demo: https://lnkd.in/eBWcPRtj

If you’re building platform tooling or running K8s in production, I’d love your feedback, especially on what you’d want scheduled next (AWS? cost scans? PR checks?).

🔌 DevOps Open Agent now supports OpenRouter 🔌

Prashant Lakhera — Fri, 26 Jun 2026 16:21:27 +0000

Most teams don’t want to be locked into a single LLM vendor, especially when you’re troubleshooting a production issue.

DevOps Open Agent now lets you route all AI analysis through OpenRouter 100+ models, one API key.

Same open-source platform. Same four agents:

✅ Kubernetes Debugging Agent

✅ AWS DevOps Agent

✅ Cloud Cost Detector

✅ PR Reviewer

Root cause analysis · PR reviews · Cloud cost insights · AWS investigations

Your model. Your choice.

Open source · Self-hostable · No vendor lock-in

🔗 GitHub: https://github.com/ideaweaver-ai/devops-open-agent

🎥 Demo: https://youtu.be/3VT8MeSLt5s

📝 Also, if you're interested in learning GenAI for DevOps, I have upcoming batches starting in July

https://ideaweaver.ai/

📌DevOps Open Agent Demo📌

Prashant Lakhera — Fri, 26 Jun 2026 02:00:53 +0000

Most DevOps tools show you data. Fewer tell you what to do next.

I recorded a full demo of DevOps Open Agent, an open-source platform for AI-powered DevOps troubleshooting.

Watch the demo:

Here’s how it works in practice:

1️⃣ One platform, four agents

✅ Kubernetes Debugging Agent — clusters, pods, networking, topology

✅ AWS DevOps Agent — EC2, Lambda, S3, VPC, CloudWatch, and more

✅ Cloud Cost Detector — unused and underutilized AWS resources

✅ PR Reviewer — DevOps-focused AI review for GitHub pull requests

2️⃣ Shared AI analysis

Pick your LLM — Anthropic, OpenAI, or Ollama — and get structured output: root cause, evidence, suggested fixes, and confidence scores.

3️⃣ Slack delivery

AI recommendations can go straight to your team’s Slack channel — no copy-paste from a dashboard.

4️⃣ Self-hostable & open source

Run it on your infra. Your data stays yours.

If you’re an SRE, platform engineer, or DevOps lead tired of jumping between 6 tools during an incident, this demo is for you.

🔗 Repo: https://github.com/ideaweaver-ai/devops-open-agent

📝 Also, if you’re interested in learning GenAI for DevOps, I have upcoming batches starting in July

https://ideaweaver.ai/

Would love your feedback in the comments.

🔥 Big update for DevOps Open Agent, Slack integration is live 🔥

Prashant Lakhera — Thu, 25 Jun 2026 16:38:34 +0000

I'm excited to share that DevOps Open Agent now supports Slack integration 🎉

When an investigation or PR review finishes, the AI analysis: root cause, recommendations, confidence, and next steps, can go straight to your favorite Slack channel.

No more:
❌ Copy-pasting RCA from the UI
❌ Losing context in DMs
Waiting for someone to check the dashboard

One platform. Four agents. One Slack channel.
✅ Kubernetes Debugging Agent
✅ AWS DevOps Agent
✅ Cloud Cost Detector
✅ PR Reviewer

Run the analysis → get the recommendation where your team already works.

Open source. Self-hostable. Vendor neutral.
🔗 https://github.com/ideaweaver-ai/devops-open-agent
📝 Also, if you're interested in learning GenAI for DevOps, I have upcoming batches starting in July
https://ideaweaver.ai/

If you're building internal DevOps tooling, I'd love your feedback.

📌 DevOps Open Agent now supports AWS Lambda and S3 📌

Prashant Lakhera — Thu, 25 Jun 2026 00:50:57 +0000

The AWS DevOps Agent can now investigate:

✅ Lambda — function discovery, timeout detection (CloudWatch + logs), invocation errors, throttles, VPC/event source issues, and AI root cause analysis with CLI fixes.

✅ S3 — bucket posture: encryption, versioning, public access, logging, notifications, and AI recommendations.

Plus dedicated Lambda and S3 troubleshooting modes so investigations stay focused, not buried in unrelated EC2 findings.

Part of DevOps Open Agent: open-source, self-hostable, AI-powered troubleshooting for Kubernetes, AWS, cloud cost, and GitHub PR review.

🔗 https://lnkd.in/giri6MF5

🔗 Also, we have a demo website available so you can get a look and feel of the interface https://lnkd.in/gMFHMSCt

📝 Also, if you're interested in learning GenAI for DevOps, I have upcoming batches starting in July

https://ideaweaver.ai/

📌 Day 2: DevOps Open Agent now supports Anthropic (Claude)📌

Prashant Lakhera — Wed, 24 Jun 2026 16:08:11 +0000

Yesterday, I released DevOps Open Agent, the first fully open-source platform built specifically for DevOps tasks

🔗https://www.linkedin.com/feed/update/urn:li:activity:7475418019783950336

❌ Not a generic chatbot.

❌ Not another coding-only agent.

A single self-hosted platform for the work DevOps teams actually do every day:

☸️ Kubernetes debugging

☁️ AWS infrastructure troubleshooting

💰 Cloud cost optimization

🔍 AI-powered GitHub PR reviews

For AI analysis, we launched with OpenAI and Ollama cloud or local, your choice.

Today, I’m happy to announce: Anthropic is now supported.

You can now run DevOps Open Agent with Claude for AI-powered investigations, root cause analysis, and PR reviews.

That gives teams more flexibility, especially for coding and infra-related tasks where Claude shines.

One platform. Four DevOps agents. Three LLM providers.

✔️ Ollama (local)

✔️ OpenAI

✔️ Anthropic ✨ new

👉 Try it: https://github.com/ideaweaver-ai/devops-open-agent

📝 Also, if you're interested in learning GenAI for DevOps, I have upcoming batches starting in July

https://ideaweaver.ai/

If you’re working in DevOps / SRE / Platform Engineering, I’d love your feedback.

🚀 Exciting news: GenAI for DevOps Engineers is back with 4 new batches starting in July!

Prashant Lakhera — Wed, 24 Jun 2026 03:37:52 +0000

If you're a DevOps, Platform, or SRE engineer wondering how actually to use Generative AI in your day-to-day work, this is built for you.

This is a hands-on, 4-week program where you build real things.

Here's what we cover:

🔹 Week 1 — Foundations: GenAI, GPUs, AI inference, running & evaluating LLMs

🔹 Week 2 — Cloud & Kubernetes: AWS Bedrock, DevOps agents, AI-powered K8s ops

🔹 Week 3 — DevOps AI Agents: PR review automation, performance troubleshooting, multi-agent systems with CrewAI

🔹 Week 4 — Advanced: MCP, RAG, fine-tuning LLMs for DevOps, production-ready AI solutions

All labs are hands-on via Google Colab no complex setup, just learn and build.

4 batches to fit your schedule:

🌅 Weekday Morning — Jul 6–29 | 7–8am PST

👉 https://lnkd.in/gi79uB4e

🌙 Weekday Evening — Jul 6–29 | 7–8pm PST

👉 https://lnkd.in/ghcfAFW6

🌄 Weekend Morning — Jul 4–26 | 7–8:30am PST

👉 https://lnkd.in/gVyDAM6h

🌇 Weekend Evening — Jul 3–26 | 7–8:30pm PST

👉 https://lnkd.in/gYXqWqMM

🔗 Detailed link: https://ideaweaver.ai/

Seats are limited; grab yours before they fill up! 🙌

If this resonates with you, share it with a DevOps or platform engineer who's been curious about AI.

📌 We don't need 10 AI agents. We need one DevOps agent 📌

Prashant Lakhera — Tue, 23 Jun 2026 16:46:48 +0000

There are hundreds of AI agents right now.

✔️ Coding agents
✔️ Research agents
✔️ Sales agents
✔️ Writing agents.

But almost none of them are built specifically for DevOps.
And honestly? I'm more than happy with one agent that actually understands DevOps work instead of five generic tools duct-taped together.
So we built DevOps Open Agent 🚀
Open source. Self-hostable. Built for SREs, platform engineers, and DevOps teams.

One platform. Four DevOps-native agents:
☸️ Kubernetes Debugging Agent
Investigate clusters, pods, networking, events, and topology, with AI root cause analysis and confidence scores.

☁️ AWS DevOps Agent
Troubleshoot EC2, VPC, security groups, load balancers, CloudTrail & more across accounts and regions. Interactive AWS topology included.

💰 Cloud Cost Detector
Find unused and underutilized AWS resources. Get savings estimates + AI-powered cost optimization recommendations.

🔍 PR Reviewer
DevOps-focused AI review for GitHub pull requests: security, reliability, and infra risks before merge.

✅ Why this matters:
Most AI for infra tools are either closed SaaS or generic chatbots.
DevOps Open Agent is purpose-built for real DevOps workflows:

➡️ investigate → analyze → recommend → track history.
⌨️ Install? One command.
git clone https://github.com/ideaweaver-ai/devops-open-agent cd devops-open-agent && chmod +x install.sh && ./install.sh
Docker Compose handles the rest.

Works with Ollama and OpenAI

👉 GitHub: https://github.com/ideaweaver-ai/devops-open-agent
If you're an SRE, DevOps engineer, or platform team lead, I'd love your feedback.

Star ⭐ the repo if this solves a problem you've felt.

🙏 If you encounter any issues, please feel free to create a GitHub issue.

🚀 Just 7 Days Left to Kick Off Our First GenAI Course for DevOps Engineers!

Prashant Lakhera — Mon, 23 Feb 2026 02:38:38 +0000

The countdown has officially started.

If you're a DevOps, Platform, or SRE engineer looking to add practical. Generative AI skills to your toolkit, this is your opportunity.

1️⃣ GenAI for DevOps Engineers

Designed specifically for engineers who want hands-on, real-world GenAI implementation — not just theory.

You’ll learn how to:

• Apply GenAI in CI/CD workflows

• Build AI-powered DevOps assistants

• Automate log analysis and troubleshooting

• Integrate LLMs into real infrastructure use cases

🔗 Curriculum: https://www.ideaweaver.ai/courses/2941959/lectures/64767222

🔗 Morning Batch:

https://www.ideaweaver.ai/l/pdp/genai-for-devops-engineers

🔗 Evening Batch:

https://www.ideaweaver.ai/l/pdp/genai-for-devops-engineers1

Seats are limited for the first cohort. Let’s build the future of DevOps with AI.