DEV Community: Israjur Rahman

How to Get Better Output from AI Tools (Without Burning Time and Tokens)

Israjur Rahman — Mon, 06 Apr 2026 14:20:14 +0000

Most engineers blame the AI when they get bad results. The real issue? The prompt.

Here's what actually works:

1. Be specific upfront
Vague prompts = vague answers.
❌ "Write a function to handle errors."
✅ "Write a Python FastAPI middleware that catches async errors and returns a structured JSON response with status code and message."

2. Use constraints
Tell the AI what not to do.
"No comments. No print statements. Use async/await with httpx, not requests."
Constraints cut bloat before it's even generated.

3. Give an example
Point it to your existing code and say "match this style." Whether you're using Claude Code, Cursor, or GitHub Copilot, letting AI read your codebase directly means it aligns with your naming conventions, patterns, and architecture, no lengthy explanation needed. If you're on a browser-based AI, just paste a snippet, same idea, same result.

4. Assign a role
"You are a senior backend engineer reviewing this API design for scalability issues."
It steers the reasoning frame and gets you a sharper, more focused review.

5. Break complex tasks apart
Don't ask AI to "build a full auth system" in one prompt.
Instead: models → routes → decorators/dependencies → pytest tests. Each step builds on the last and errors are easier to catch.

6. Refine, don't regenerate
Something's off? Don't restart. Say:
"This Python function is returning None instead of the parsed JSON, debug just this function, don't touch the rest."
Targeted edits save tokens and preserve what's already working.

7. Control output length
"Give me 3 approaches to this caching problem, one paragraph each."
Longer output ≠ better output. It just takes more time to read and review.

8. Know when AI can mislead you
Designing system architecture, making security-critical decisions, or estimating performance at scale, AI can sound very confident and still be completely wrong. Always validate its output with your own judgment and domain knowledge.

The core principle?

AI won't fix a bad brief. The quality of your output is directly proportional to the clarity of your input.

Some basic docker commands

Israjur Rahman — Fri, 25 Jun 2021 09:56:24 +0000

Some basic docker commands that will come handy when working with docker.

To build a docker project

docker-compose build

To build & run docker project

docker-compose up

To stop a docker project

docker-compose down

To show docker image lists

docker images

To remove a docker image

docker rmi <IMAGE ID>

To show only running containers

docker ps

To show all containers

docker ps -a

To show the latest created container

docker ps -l

To display total file size of containers

docker ps -s

To remove a container

docker rm <ID or NAME>

To run django command in docker project

docker exec -it <CONTAINER ID> python manage.py migrate

How to implement asynchronous socket status using django channels

Israjur Rahman — Sat, 23 Jan 2021 09:04:16 +0000

While developing an analytical project using django we came across an issue where we have to show status when our users start analysis. We decided to implement this using django channels. There are many tutorials about django channels on the internet but all of them are a little complex or using django channels to build a chat application. So I decided to write this article where I will show how to send asynchronous status to users using django channels.

I am assuming you know at least the basics of django framework and how django works, so I am not going to show how to create a django project and django apps. I will directly start from implementing django channels.

First we will start a project channelproj and create an app notifier.

We need to install channels and channels-redis using pip :

pip install channels channels-redis

We need to install redis-server to out system:

sudo apt update
sudo apt install redis-server

Check redis-server status:

sudo systemctl status redis

Here we can see redis server is active and running on port 6379

Then open django channelproj/settings.py file and add channels and notifier to INSTALLED_APPS.

INSTALLED_APPS = [
    'django.contrib.admin',
    'django.contrib.auth',
    'django.contrib.contenttypes',
    'django.contrib.sessions',
    'django.contrib.messages',
    'django.contrib.staticfiles',
    'channels', # Here we added channels 
    'notifier' # we will use django channels in this app
]

We also need to add this lines in settings.py file:

ASGI_APPLICATION = 'channelproj.asgi.application'

CHANNEL_LAYERS = {
    "default": {
        "BACKEND": "channels_redis.core.RedisChannelLayer",
        "CONFIG": {
            "hosts": [("127.0.0.1", 6379)],
        },
    },
}

To know more about channel layer click here

Now we will create a file notifier/consumers.py and add the following class

from channels.generic.websocket import AsyncJsonWebsocketConsumer

class StatusConsumer(AsyncJsonWebsocketConsumer):
    room_group_name = 'notify'

    async def connect(self):
        await self.channel_layer.group_add(
            self.room_group_name,
            self.channel_name
        )
        await self.accept()

    async def disconnect(self, code):
        await self.channel_layer.group_discard(
            self.room_group_name, 
            self.channel_layer
        )

    async def status_notifier(self, event):
        await self.send_json(event)

Here we are using AsyncJsonWebsocketConsumer and creating a group named notify and we have defined a function called status_notifier which will send all the events via socket url as json format.

We will create another file notifier/routing.py and add our socket urls

from django.urls import path

from .consumers import StatusConsumer

ws_urlpatterns = [
    path('ws/status/', StatusConsumer.as_asgi())
]

Now open channelproj/asgi.py file and edit the file as below:

from channels.routing import ProtocolTypeRouter, URLRouter
from django.core.asgi import get_asgi_application

from notifier.routing import ws_urlpatterns

os.environ.setdefault('DJANGO_SETTINGS_MODULE',
'channelproj.settings')

application = ProtocolTypeRouter({
    "http": get_asgi_application(),
    "websocket": URLRouter(
        ws_urlpatterns
    )
})

Here we have imported ws_urlpatterns from notifier/routing.py added ProtocolTypeRouter and defined our websocket urls using URLRouter

Now we will create a template notifier/templates/home.html in our notifier app where we will add a form with an input field.

  {% load static %}
  <html>
    <head>
      <title>Django Channels Status</title>
    </head>
    <body>
      <h1>Check Status</h1>

    <form action="{% url 'staus' %}" method="post">
      {% csrf_token %}
      <input type="text" placeholder="enter anything" name="TB_sample" id="TB_sample"/><br>
      <input type="submit" value="submit">
    </form>

  </body>
  </html>

Now we will add a two function in notifier/views.py

from django.shortcuts import render, redirect
from asgiref.sync import async_to_sync
from channels.layers import get_channel_layer
import time


def index(request):
    return render(request, 'home.html')

def status_form(request):
    if request.method =='POST':
        num = int(request.POST['TB_sample'])
        progress = 10
        for i in range(num):
            room_group_name = f'notify'
            channel_layer = get_channel_layer()
            async_to_sync(channel_layer.group_send)(
                room_group_name, {
                    "type": "status.notifier",
                    "data": progress
                }
            )
            message = "Status Running"
            progress += 10
            time.sleep(1)

    context = {'message': message}
    return render(request, 'home.html', context)

Here we have defined a function status_form which will be called when a post request happens, it will get the post value in num variable and we have declared a variable called progress with value 10, we have defined a loop which will run and add 10 every time loop runs to progress and sleep for 1 second. We have added channel_layer = get_channel_layer() line to get the channel layer and async_to_sync method to send the data asynchronously to room_group_name.

Note: Here by declaring "type": "status.notifier" we are calling status_notifier function we wrote in notifier/consumers.py file.

Also edit channelproj/urls.py file like this:

from notifier.views import index,status_form

urlpatterns = [
    path('', index),
    path('status/', status_form, name="status"),
    path('admin/', admin.site.urls),
]

Now if we run the project using python manage.py runserver and open http://127.0.0.1:8000/ url in our browser it will look something like this

To check socket status we will use a chrome extension called Simple WebSocket Client. Add the extension and click the extension icon in chrome and insert our web socket url ws://127.0.0.1:8000/ws/status/ and click open, connection should be established.

Now go to project tab and insert 10 in input filed and submit the form and move to Simple WebSocket Client tab. We will see the status we are sending to socket appearing every second until the loop completes.

I have uploaded a sample project to github you can get here.

Thank you for reading!

Convert Microsoft SQL Server Database Tables to CSV

Israjur Rahman — Fri, 22 Jan 2021 08:46:45 +0000

We can convert and save Microsoft SQL Server database tables to CSV using python. I've recently written a python script to automate a process to convert Microsoft SQL Server database tables to CSV format and save it to drive so that I can use it later for other purposes. I've used pyodbc and pandas library to automate this process.

Firstly we have to establish a connection to Microsoft SQL Server database using pyodbc library.

import pyodbc 

conn = pyodbc.connect('Driver={SQL Server};'
                      'Server=ISRAJ\SQLEXPRESS;'
                      'Database=TEST_DB;'
                      'Trusted_Connection=yes;')

cursor = conn.cursor()

Here we have provided server name, database name and if it has any username and password we have to add the credentials too.

After that we will query table names so that we can see the tables we have and can select the table name we want to convert to CSV.

sql = "SELECT TABLE_NAME FROM TEST_DB.INFORMATION_SCHEMA.TABLES WHERE TABLE_TYPE = 'BASE TABLE'"

cursor.execute(sql)

We have written a query to select all the tables from database and by using cursor.execute we are executing the query.

Now we can loop through the table names:

for (table_name,) in cursor:
    print(table_name)

Output will be something like this:

Table_One
Table_Eight
Table_Two
Table_Three
Table_Four
Table_Five
Table_Six
Table_Seven
Table_Nine

Now we will use pandas to get the specific table data to dataframe.

import pandas.io.sql as psql

sql = "SELECT * FROM TEST_DB.dbo.Table_One"
df = psql.read_sql(sql, conn)

Here we have written another query to select all from Table_One and psql.read_sql will execute the query and convert the data into dataframe.

Now we will use pandas to_csv method to save the dataframe as csv.

df.to_csv("table_one_in_csv.csv", index=False)

Thanks For Reading!

If you want to test or get the code here I’ve added a simple command line tool to convert Microsoft SQL Server database tables to CSV.

https://github.com/IsrajurRahman/mssql2csv