DEV Community: Amrendra N Mishra

I Control My Mac with Voice — Say Hey Jarvis and It Does Everything

Amrendra N Mishra — Sun, 12 Jul 2026 12:54:16 +0000

I built a voice assistant that controls 45 AI tools. I say "Hey Jarvis" and it executes.

What It Does

Command	Action
"generate content"	Creates YouTube scripts for 9 channels
"research quantum computing"	Deep research via Tavily + AI
"write email about meeting"	Drafts email, copies to clipboard
"start focus"	Starts Pomodoro + blocks apps
"code review"	Reviews git diff with AI
"summarize"	Summarizes clipboard content
"find file tax PDF"	Natural language file search

Architecture

Mic → Whisper (offline) → Intent Classify → Router → Ollama → say (TTS)

Key Features

Offline speech (Whisper local)
Wake word: "Hey Jarvis"
Global hotkey: Ctrl+Space
Command chaining: "research AI then write blog"
Memory across conversations
Hindi + English

Setup

brew install portaudio
pip install SpeechRecognition pyaudio openai-whisper
python voice_commander_pro.py

🔗 github.com/amrendramishra/ai-tools
🌐 amrendranmishra.dev

Building a Multi-Agent AI System from Scratch (No Frameworks)

Amrendra N Mishra — Sun, 12 Jul 2026 12:53:28 +0000

You dont need CrewAI or AutoGen. Heres how to build a multi-agent pipeline in pure Python.

The Concept

5 AI agents collaborate in a pipeline:

Researcher → Writer → Editor → Reviewer → Publisher

Each agent has a role, a system prompt, and passes output to the next.

Implementation

class Agent:
    def __init__(self, name, role, system_prompt):
        self.name = name
        self.role = role
        self.system_prompt = system_prompt

    def process(self, input_text):
        """Call LLM with system prompt + input"""
        import requests
        response = requests.post(
            "http://localhost:11434/api/generate",
            json={
                "model": "llama3.2",
                "prompt": f"{self.system_prompt}\n\nInput: {input_text}",
                "stream": False
            }
        )
        return response.json()["response"]


class Pipeline:
    def __init__(self, agents):
        self.agents = agents

    def run(self, topic):
        output = topic
        for agent in self.agents:
            print(f"🤖 {agent.name} working...")
            output = agent.process(output)
            print(f"   ✅ Done ({len(output)} chars)")
        return output


# Define agents
agents = [
    Agent("Researcher", "research",
          "Find 5 key facts about the topic. Be concise."),
    Agent("Writer", "write",
          "Write a 500-word article from the research."),
    Agent("Editor", "edit",
          "Improve clarity, fix grammar, make it engaging."),
    Agent("Reviewer", "review",
          "Score 1-10. List what works and what needs fixing."),
    Agent("Publisher", "publish",
          "Format as markdown with title, headings, and CTA."),
]

# Run
pipeline = Pipeline(agents)
result = pipeline.run("The Future of Local AI")
print(result)

Why No Framework?

Simpler: 50 lines vs 500 lines with CrewAI
Debuggable: You can see exactly whats happening
Flexible: Add/remove agents easily
No dependencies: Just requests library
Works with any LLM: Ollama, OpenAI, Groq, anything

Pro Tips

Different models per agent: Use codellama for code review, llama3.2 for writing
Temperature: Lower for Researcher (factual), higher for Writer (creative)
Fallback: If one model fails, try another
Logging: Save each agents output for debugging

Use Cases

Blog post pipeline
Video script generation
Social media content
Research reports
Code review chains

🔗 Full implementation: github.com/amrendramishra/ai-tools
🌐 amrendranmishra.dev

Building AI tools daily. Follow for more.

How I Automated 9 YouTube Channels at ₹0/month

Amrendra N Mishra — Sun, 12 Jul 2026 12:53:27 +0000

I run 9 YouTube channels that post daily. I dont touch any of them. Total cost: ₹0.

The Architecture

Topic CSV → AI Script (Gemini/Ollama) → Google TTS → MoviePy Video → YouTube Upload

All scheduled via Mac LaunchAgents. Each channel has its own plist file that triggers at a specific time.

The 9 Channels

Channel	Niche	Time
Gyaan in 5	Knowledge	6:15 PM
Tech in 5 Hindi	Tech	11:30 AM
Money in 5	Finance	9:30 AM
Superhuman 60s	Productivity	2:30 PM
Horror Ki Kahani	Horror	9:30 PM
Pyaar Ka Psychology	Relationships	8:30 PM
Zeheela Sach	Dark Facts	7:30 PM
Apna Haq	Legal Rights	12:30 PM
Agar Aisa Ho Toh	Hypothetical	1:30 PM

The Stack (All Free)

Script generation: Gemini API (free tier, 4-key rotation) + Ollama fallback
Voice: Google Cloud TTS (Wavenet, free 4M chars/month)
Images: Pexels API (free stock images)
Video: MoviePy (Python library, 1080x1920 vertical)
Upload: YouTube Data API v3 (free)
Scheduling: macOS LaunchAgents (built-in, free)

Key Trick: 4-Key Rotation

import random

API_KEYS = [
    "key_1_from_project_A",
    "key_2_from_project_B",
    "key_3_from_project_C",
    "key_4_from_project_D",
]

def get_api_key():
    return random.choice(API_KEYS)

This avoids hitting rate limits on any single key.

Monthly Cost Breakdown

Service	Cost
Google TTS	₹0 (free tier)
Pexels	₹0 (free API)
YouTube API	₹0
Ollama	₹0 (local)
Mac LaunchAgents	₹0 (built-in)
Total	₹0

The Result

9 channels × 1 video/day = 270 videos/month. Zero effort after initial setup.

🔗 Full automation code: github.com/amrendramishra/ai-tools
🌐 Portfolio: amrendranmishra.dev

Follow for daily AI automation content.

5 AI Tools You Can Run Locally for Free in 2026

Amrendra N Mishra — Sun, 12 Jul 2026 12:28:54 +0000

Are you paying for AI APIs? Stop. Here are 5 tools that run on your laptop for free.

1. Ollama

Run any LLM locally:

brew install ollama
ollama pull llama3.2

Now you have GPT-level AI on your machine. No API key needed.

2. ChromaDB

Free vector database. Store embeddings locally:

import chromadb
client = chromadb.Client()
collection = client.create_collection("docs")
collection.add(documents=["AI is amazing"], ids=["1"])
results = collection.query(query_texts=["tell me about AI"])

3. Whisper

Offline speech recognition by OpenAI:

import whisper
model = whisper.load_model("base")
result = model.transcribe("audio.mp3")
print(result["text"])

4. LangChain

Build AI apps that connect to your data:

from langchain_community.llms import Ollama
llm = Ollama(model="llama3.2")
print(llm.invoke("Explain Docker in 2 lines"))

5. MCP (Model Context Protocol)

Connect AI to any tool — files, GitHub, databases. The USB-C of AI.

My Setup

I built 45 tools using these. All free. All local. All open source.

🔗 github.com/amrendramishra/ai-tools
🌐 amrendranmishra.dev

Follow for daily AI engineering content.