Akansha Yadav | AI & Python | Substack

Akansha Yadav | AI & Python

Learning AI by building everything from scratch. I write about:

• Neural Networks
• LLMs
• Python for AI
• Real AI projects

Explaining complex AI simply.

Over 1,000 subscribers

Akansha Yadav | AI & Python

5 misconceptions about data analytics (and what's actually true):

❌ The more sophisticated the tool, the better the analyst
✅ Many analysts do their jobs with "basic" tools like Excel
❌ You're just there to crunch the numbers
✅ You need to be able to tell a story with the data
❌ You need super advanced math skills

TikTok’s parent company just dropped DeerFlow 2.0, an open-source framework for building AI agents — and it’s currently #1 on GitHub Trending.

Here’s the gist of what it does:

🔹 Parallel task handling: Breaks large problems into chunks and lets multiple agents tackle them at once
🔹 Cross-session memory: Keeps track of your previous workflow for better context

Why Most People Quit Learning Python (And How to Avoid It)

They fail because they try to learn everything at once.

10 End-to-End Data Analyst Projects with GitHub Links to Build a Job-Winning Portfolio

  1. End-to-End Retail Analytics Dashboard Project Using SQL and Power BI
    GitHub: github.com/SaiKiran-Dat…
  2. Customer Churn Prediction and Dashboard Project for Data Analysts

10 SQL Projects You Can Start Today (With Datasets)

  1. Sales Performance Tracker
    Perfect for learning KPIs, revenue trends, and top products.
    kaggle.com/datasets/kya…

Top Remote Job Websites

  1. jobspresso.co
  2. virtualvocations.com
  3. stackoverflow.com/jobs
  4. outsourcely.com

Everybody Wants to Use ChatGPT. Very Few Understand How It Actually Works.

🔥 Google Colab has added the option of retraining 500+ open-source neural networks
Unsloth has released a convenient notebook for configuring models.

Instructions:

  1. Open the page in Colab: colab.research.google.c…
  2. Run the blocks and the Unsloth Studio itself.
  3. Select a model and a dataset.
  4. Click "Start Training" and monitor the progress in real time.
  5. Everything is ready - you can immediately compare the regular and fine-tuned versions of the model in the chat.