Ashish Gupta’s Post

When people talk about AI testing, they often jump straight into complex tools. But one thing I’ve started realizing: Python is the real foundation behind it. While exploring AI testing, I began focusing on Python basics again — and it changed my perspective 🔹 Handling API responses using Python 🔹 Validating data instead of exact outputs 🔹 Writing flexible assertions for unpredictable results 🔹 Working with libraries like Requests & PyTest Because in AI systems: Outputs are not always the same Traditional “expected vs actual” doesn’t always work That’s where Python helps — it gives the flexibility to analyze, validate, and adapt test logic. I’m still at the beginning of this journey, but one thing is clear: Strong Python skills are essential for anyone moving into AI testing. Next, I’m exploring: How to validate AI/ML model responses Data-driven testing approaches Learning step by step. How are you using Python in your testing journey? #Python #AITesting #MachineLearning #QA #AutomationTesting #SoftwareTesting #Learning #TechJourney

AI itna fast improve kyun ho raha hai? Answer: Python libraries. 🐍☠️ Python khud fastest language nahi hai, lekin ecosystem unbeatable hai. Why it works: • NumPy → fast computations • Pandas → easy data handling • PyTorch / TensorFlow → deep learning in few lines • Hugging Face → ready-to-use models • LangChain → AI agents fast build Python isn’t fast. It makes fast systems usable like C++. Result: Ideas → Code → Test → Iterate Now happens in days, not months. That’s why AI is moving so fast.

AI Beyond the Hype | Part 8: Vector Databases “What is Python used for?” “Is python dangerous?” Same word. Completely different meaning. 👉 In one case → Python = programming language 🧑💻 👉 In another → python = reptile 🐍 We can’t store every possible variation or phrasing. Traditional search fails here because it works on exact match, not meaning. This is where semantic search (search based on meaning) comes in — and that’s where vector databases play a key role. ## 🧠 What is a Vector Database? A vector DB stores data as embeddings (numbers) instead of plain text, so it can search based on meaning. ## 🔢 How data is generated and stored Text → tokens → embeddings Example: “Python is used for backend development” → [0.12, -0.45, 0.78, …] “Python is a dangerous reptile” → [-0.33, 0.91, -0.12, …] These numbers capture meaning, not just words. ## 🔍 How search happens User query → embedding Example: “Python coding” → vector “Is python poisonous” → vector Then system finds vectors that are closest in meaning (not exact match). This is semantic search. ## ⚡ How search is optimized Searching millions of vectors directly is slow. So vector DBs use indexing (ANN – Approximate Nearest Neighbors) and sometimes hashing/partitioning to find nearest vectors quickly. ## 🧩 How prompt-based retrieval works 1. Query → embedding 2. Retrieve relevant chunks 3. Add to prompt 4. LLM generates answer → This is how RAG works internally. ## 🚨 Reality check Vector DB doesn’t understand meaning. It just finds patterns that are mathematically close. ## ⚠️ Challenges Similar ≠ correct Bad embeddings → bad retrieval Needs tuning (top-k, thresholds) Scaling & latency trade-offs ## 💡 Takeaway 👉 “Vector DB doesn’t search words — it searches meaning.” Funny how things work — what felt pointless in school is now the backbone of AI systems

Built another Python web scraping project while learning data collection for AI/ML. This time I created a scraper that collects book data from an online catalogue. Repo - https://lnkd.in/gej-ZwFG The scraper: • Extracts book titles • Scrapes prices and ratings • Automatically navigates through all pages • Stores the dataset in JSON format While building it I practiced concepts like HTML parsing, pagination scraping, and handling relative URLs. Learning web scraping step by step and building small projects along the way. #Python #WebScraping #BuildInPublic #Learning #AI

"Should I use Python or Elixir for AI?" It's like asking whether a kitchen needs a chef or a maître d'. You need both — just not for the same things. Here's the breakdown I wish existed when I started. Python is the undisputed home of AI research. Every breakthrough model ships a Python reference implementation first. Hugging Face has 500k+ models. The data science ecosystem —NumPy, Pandas, PyTorch — is a decade deep. There's no replacing it for training and experimentation. But then comes production. Suddenly you need to: → Handle thousands of concurrent API requests → Stream tokens live to real users → Keep a fleet of AI agents running 24/7 without crashing → Deploy new model versions without dropping live traffic This is where Python starts to show its weakness and where Elixir, built on the battle-tested Erlang VM, excels. Over the next 6 posts, I'll walk through practical, code-level comparisons across concurrency, inference, training, real-time streaming, agentic workflows, and fault tolerance. No tribalism. Just the right tool for each layer. 📌 Save this post — the series starts April 16th, 2026. #Elixir #Python #MachineLearning #SoftwareEngineering #AI

Teaser for an upcoming series of posts discussing Python, Elixir, and AI tools. Check back Thursday for the first post!

🐛 A subtle Python bug that corrupted an entire AI-generated document I was using a LangGraph RAG agent to generate content for multiple document sections in parallel — each section with its own question. But the final document had duplicate content across sections. Different sections, same output. My first instinct was LangGraph. I debugged the agent, checked configurations, changed settings. Nothing worked. Then I printed the memory address of the dictionary being passed inside each thread. Every thread. Same memory address. That was the moment everything clicked. The threads weren't working with their own data — they were all pointing to the same dictionary object in memory. Each thread was overwriting the question, so every thread ended up running with the same question. Fix: One line — copy.deepcopy() inside the thread, before passing the dict to the agent. Every thread got its own independent copy. Unique questions. Unique outputs. Problem solved. Lesson: When parallel outputs look suspiciously similar — before blaming the AI, check your memory. #Python #LangGraph #RAG #Debugging #Multithreading #AI #LLM #ProblemSolving #MachineLearning

10 years ago, Python was "that scripting language." Today, it's the backbone of the AI/ML revolution. And I don't think most people appreciate how fast that shift happened. Here's what changed: NumPy gave us fast numerical computing in Python. Then came pandas, then scikit-learn. Each library solved a real problem, and the ecosystem snowballed. Then PyTorch and TensorFlow arrived. Suddenly, Python wasn't just analyzing data. It was training neural networks that could see, read, and generate. Now with LLMs? Python is the default language for every AI prototype, pipeline, and production system being built right now. But here's what this means for us as Python developers: The bar has shifted. Writing clean, functional code is still the foundation. But today's Python developer is also expected to understand data pipelines, model evaluation, vector databases, and API integrations with AI services. It's a lot. And it's only accelerating. My take: you don't need to become a data scientist or ML researcher. But you do need enough fluency to build around these systems to connect the pieces, ask the right questions, and deliver products that actually use AI meaningfully. The opportunity for Python developers right now is enormous. The question is whether we're keeping up with it. Are you upskilling in data/ML or staying focused on your lane? Curious where others are drawing the line. #Python #MachineLearning #DataScience #C2C #C2H #ArtificialIntelligence #SoftwareEngineering

Day 24 of my 60-Day Python + AI Roadmap. 🚀 Every program will face unexpected inputs. Every AI model will receive bad data. The question is — does your code crash or handle it? Today I learned how to make Python bulletproof. 🛡️ 🔥 OPINION — Agree or Disagree? "An AI model that crashes on bad input is useless in production. Exception handling isn't optional — it's what separates a script from a real product." Comment AGREE 🟢 or DISAGREE 🔴! 🧠 GUESS THE OUTPUT — Before you scroll! try:  x = int("abc") except ValueError:  print("Invalid number") else:  print("Success") finally:  print("Done") ⚠️ except + else + finally — all 3 together! Answer at 50 comments 🎯 ━━━━━━━━━━━━━━━━ Exception Handling — Key Concepts ━━━━━━━━━━━━━━━━ 🔴 try → risky code goes here 🟡 except → what to do if it fails 🟢 else → runs ONLY if no error occurred ⭐ finally → ALWAYS runs (error or not) 🤖 AI use — real example: try:  prediction = model.predict(data) except ValueError:  print("Invalid input shape") finally:  log.close() ✅ Common exceptions to know: ValueError → wrong value type TypeError → wrong data type ZeroDivisionError → divide by zero FileNotFoundError → file missing 💡 Analogy: try → Trying something risky 🪂 except → Parachute opens if it fails else → Landing perfectly ✅ finally → Always pack your bag back 🎒 🚨 Golden Rules: ❌ Never use bare except: — catches everything silently! ✅ Always catch specific exceptions ✅ Keep try block as small as possible --- 👆 What does the code above print? Drop answer + AGREE 🟢 / DISAGREE 🔴 below! 👇 On a learning journey? Drop your day number! 🤝 💾 Save · ♻️ Repost #60DayChallenge #Python #ExceptionHandling #LearnPython #PythonForAI #MachineLearning #AILearning #100DaysOfCode #LearningInPublic #BuildInPublic #DataScience #CodeNewbie

🚀 Quick Reminder: Python Strings & Methods Today I revised one of the most important basics in Python — Strings & their Methods 🐍 🔹 A string is simply a sequence of characters inside quotes. Example: "Hello World" 💡 Must-Know String Methods: ✅ Case Conversion upper(), lower(), title() ✅ Searching find(), index(), count() ✅ Modify replace() ✅ Remove Spaces strip(), lstrip(), rstrip() ✅ Join & Split split(), " ".join() ✅ Check Methods isalpha(), isdigit(), isalnum() ✅ Other Useful Ones startswith(), endswith(), len() 🧠 Mini Practice: Count vowels Check palindrome Remove duplicates Find character frequency ⚡ Quick Tip: Strings are immutable, which means they cannot be changed directly. 📌 Mastering strings is very important for data cleaning, NLP, and AI projects. #Python #Coding #LearningJourney #100DaysOfCode #AI #Programming #Students #PythonBasics