Pooja Sahu’s Post

How to capture the latest news, automatically Ever feel like “staying updated” is a second full-time job? I’m testing a simple workflow to reduce that load. Using Python + Obsidian, I periodically check a curated set of sources and turn updates into searchable notes. I’m not sharing the collected content—only the collection method. Early results: less manual searching, fewer rabbit holes, and better traceability. The focus is on inputs + tagging + retrieval (so it’s useful later, not just “saved”). Also important: being mindful of TOS/robots and keeping requests lightweight. Still tuning filters to cut noise and capture only meaningful changes. #Productivity #Automation #Python #Obsidian #KnowledgeManagement

The "No API" Excuse is Dead 🚫 "Sorry, that system is too old. It doesn't have an API." I hear this in meetings all the time. It usually means a project is dead in the water. But here is the truth: If it has a user interface, it has an API. It's just an API designed for humans, not computers. The Inputs: Text boxes and buttons. The Outputs: Screen text and tables. With Python and Playwright, we can turn that "Legacy User Interface" into a structured, programmable API. We don't need to wait for a digital transformation project. We can build the bridge today. #LegacySystems #API #Python #Playwright #DigitalTransformation Truly yours Bot.

We all hit those moments. You're debugging a complex function, and a simple, repeated calculation keeps tripping you up. It feels like you're writing the same logic over and over, tangled in nested loops that are hard to trace. This often happens because we're trying to solve a problem by breaking it down into smaller, identical versions of itself. We're thinking about "what's the answer for this input, and how does it relate to the answer for a slightly smaller input?" The fix is to embrace recursion. Think about calculating factorial. 5! is 5 4!. And 4! is 4 3!, and so on, until you reach the simplest case: 1! is just 1. In Python, this looks elegant: def factorial(n): if n == 1: return 1 else: return n * factorial(n - 1) The function calls itself with a smaller input until it hits the base case. This approach leads to cleaner, more readable code. It simplifies complex problems into understandable, self-similar steps. Less code to write. Fewer bugs to find. Easier to reason about. #Python #Recursion #SoftwareEngineering #CodingTips #SeniorEngineer

Open-source release: Document Text Extractor (Python) I’ve been working on document-text-extractor, a modular and well-tested Python package for extracting text from PDFs, scanned documents, and images — with OCR fallback. The tool provides: - A CLI for automation and batch processing. - A Streamlit GUI for quick inspection and demos. - A reusable Python package you can integrate directly into your projects. - Efficient memory management, designed with large documents and pipelines in mind. It’s particularly useful for RAG ingestion pipelines, where clean, reliable text extraction is a prerequisite for chunking, embeddings, and LLM workflows. I’m sharing it publicly to get real engineering feedback, especially around: - performance and accuracy vs existing tools - multi-language OCR strategies - integration patterns with LangChain or LlamaIndex Repo 👉 https://lnkd.in/ddr_UQ7y Feedback, issues, and contributions are very welcome. #opensource #python #rag #llm #ocr #engineering #machinelearning #devtools

Stop Using Python Multiprocessing Like It’s 2015, This Is the 2026 Way I thought I was being clever using Process(), start() and join() everywhere. Turns out I was doing multiprocessing the old way. Here’s the shift that finally worked: Old style: Manually creating processes, managing workers, printing results, no real error handling. It doesn’t scale. Modern style: Use process pools, not manual processes. Let Python manage the workers. Use map() / starmap() for clean return values. For real projects → ProcessPoolExecutor. Add progress with imap_unordered(). Set a smart chunksize so overhead doesn’t kill performance. Don’t use multiprocessing for: - Tiny datasets - Fast tasks - I/O-bound work (use threads / asyncio instead) Real lesson: Multiprocessing isn’t about creating processes. It’s about distributing CPU work efficiently. If you’re still doing Process() + start() + join() manually… you’re working too hard. What’s your go-to setup for speeding up Python now? #Python #Multiprocessing #Performance #Concurrency #SoftwareEngineering

Tech Tactic Today I implemented Python Scripts in my workflow and had a lightbulb moment. Think of scripts like tiny robot assistants. Instead of manually doing repetitive tasks, I write a few lines of code that do it for me. Example: I used to spend 2 hours every week updating my analytics spreadsheets. Now? A 20-line Python script does it in 30 seconds. It's like building a machine that builds other machines. Once you write a script, it's yours forever. No more manual grunt work. I'm kicking myself for not doing this sooner. Could've saved hundreds of hours this year alone. What's one repetitive task in your business that you wish you could automate? Lina V. #Automation #ProductivityHacks #SoloFounder #BuildInPublic #Python

🚀 Day 5/50 – DSA with Python Challenge Today I focused on Python String Data Structure and practiced many important concepts. 📌 What I learned & practiced today: 🔹 String creation & escape characters Single quotes, double quotes Escape sequences (\n, \\, \') 🔹 String operations Indexing (positive & negative) String comparison (<, >, ==, !=) ASCII values using ord() and chr() 🔹 String formatting % formatting format() method f-strings 🔹 String methods find(), index() startswith(), endswith() split(), join() strip(), lstrip(), rstrip() Membership operator (in) 🔹 Pattern searching in strings Finding all occurrences using find() with loop 🔹 Palindrome checking Using two-pointer technique Using slicing ([::-1]) 🔹 Anagram checking Using sorting method Using character frequency (ASCII count – optimized approach) 🧠 Key takeaway: Understanding strings deeply helps in solving many DSA problems efficiently and builds strong problem-solving fundamentals. 📈 Consistency over motivation — one day at a time. #Day5 #DSAWithPython #PythonProgramming #StringDataStructure #LearningInPublic #50DaysChallenge #ProblemSolving #CodingJourney

LeetCode Progress | 258. Add Digits (Python) Today I solved “Add Digits” on LeetCode. Problem: Given an integer num, repeatedly add its digits until the result has only one digit, and return it. My approach: I used an iterative digit-sum process. -- Repeatedly extracted digits using modulo and division -- Summed the digits until the number became a single digit -- Returned the final value Optimal approach (O(1) without loops): This problem follows the mathematical concept of a Digital Root. The result can be found using: -- If num == 0, return 0 -- Otherwise, return 1 + (num - 1) % 9 This avoids iteration entirely and runs in constant time. What I learned: -- Some problems that look iterative have hidden mathematical patterns -- Recognizing number properties (like digital roots) can lead to O(1) solutions -- Always look for a formula when repeated digit operations are involved #leetcode #python #dsa #datastructures #algorithms #coding #programming #problemSolving #softwareengineering #computerscience #interviewprep #codinginterview #100daysofcode #pythonprogramming

𝗣𝘆𝘁𝗵𝗼𝗻 𝗗𝗮𝗶𝗹𝘆 𝗖𝗵𝗮𝗹𝗹𝗲𝗻𝗴𝗲 | 𝗛𝗮𝗰𝗸𝗲𝗿𝗥𝗮𝗻𝗸 – 𝗰𝗼𝗹𝗹𝗲𝗰𝘁𝗶𝗼𝗻𝘀.𝗖𝗼𝘂𝗻𝘁𝗲𝗿() | 𝗗𝗮𝘆 𝟮𝟯 This Python tool replaces an entire frequency loop. Day 23 of my Python Daily Challenge 🚀 𝗧𝗼𝗱𝗮𝘆’𝘀 𝗽𝗿𝗼𝗯𝗹𝗲𝗺 𝗳𝗲𝗹𝘁 𝗽𝗿𝗮𝗰𝘁𝗶𝗰𝗮𝗹: 👉 Track shoe sizes 👉 Handle multiple customers 👉 Calculate total earnings 𝘠𝘰𝘶 𝘤𝘰𝘶𝘭𝘥 𝘴𝘰𝘭𝘷𝘦 𝘵𝘩𝘪𝘴 𝘸𝘪𝘵𝘩 𝘭𝘰𝘰𝘱𝘴 𝘢𝘯𝘥 𝘥𝘪𝘤𝘵𝘪𝘰𝘯𝘢𝘳𝘪𝘦𝘴… or you could think like Python 👇 𝗧𝗵𝗲 𝗿𝗲𝗮𝗹 𝗶𝗻𝘁𝗲𝗿𝘃𝗶𝗲𝘄 𝗹𝗲𝘀𝘀𝗼𝗻: • Counter handles frequency cleanly • Updates reflect real-world inventory logic • Less code ≠ less clarity 💡 𝗜𝗻𝘁𝗲𝗿𝘃𝗶𝗲𝘄 𝗽𝗮𝘁𝘁𝗲𝗿𝗻 𝗳𝗿𝗼𝗺 𝗗𝗮𝘆 𝟮𝟯: Strong candidates know when to stop reinventing logic and start using the right data structure. That’s how clean, scalable solutions are written. Have you used Counter in real problems before? 👇 #Python #HackerRank #DailyCoding #ProblemSolving #InterviewPrep #LearnInPublic #Consistency

Problem: Manual data cleaning every month 😬 If you do the same steps repeatedly, automate it. One Python trick I use often: 👉 pandas apply + functions for standardization Example: Different date formats Inconsistent department names Instead of fixing it manually: df['dept'] = df['dept'].str.strip().str.upper() 5 seconds of code vs 30 minutes of manual work. Python isn’t just for ML. It’s for saving analyst time. What’s the most repetitive task you still do manually?