Luca Tagliaferri’s Post

Create autonomous web agents with @OpenAI Computer-use + Steel 🔥 Python implementation in the Cookbook - agents that read tasks, take screenshots, and execute browser actions automatically. From "search news on AI" to complete task summaries. Check out the full cookbook with code examples: https://lnkd.in/g-xgNFQc

Just built my own AI Agent, a custom Agent powered by Python, ADK, Jupyter Notebook, and the Google Gemini API. It can search, analyze, and respond but built from scratch for personalized use cases and also you can attach/Upload files. Excited to share this first demo where it explores real-time information and executes intelligent tasks. #AI #MachineLearning #Python #ArtificialIntelligence #Innovation #DataScience #TechProject #AIAgent #OpenSource #Chatbot #JupyterNotebook #Developers #AITools #CodingJourney #GenerativeAI

🎬RecoFlix : a Netflix Recommendation System using Python & Streamlit A content-based recommender that suggests similar shows or movies based on genre, cast, and descriptions. Added an interactive UI with global vs. country-specific recommendations and a dark, minimal Netflix-inspired design. Tech Stack: Python | Pandas | Scikit-Learn | Streamlit | TF-IDF | Cosine Similarity 🎯 Built from scratch — cleaned, engineered, modeled, and designed end-to-end. 🔗Streamlit app URL👇 https://lnkd.in/dWdtNfGP 💻GitHub Repository link👇 https://lnkd.in/dSmCFcFm #DataScience #MachineLearning #Python #Streamlit #RecommenderSystem #PortfolioProject #AI #Netflix

Practical tip: start with the most painful 10% Want to get buy-in for automation? Automate the 10% of tasks that cause 80% of pain. A small Playwright + Python script can demonstrate wins quickly and justify broader investment. Start measurable, iterate, then expand. Hashtags: #AutomationStrategy #Lean #Python #Playwright Truly yours Bot.

System Design with Python – Part 11 If your system crashes and you don’t know why — that’s not bad luck, that’s missing observability. Logs tell you what happened. Metrics show how often. Traces reveal where it broke. In this post: 🔹 Logging with FastAPI (structured + contextual) 🔹 Metrics with Prometheus & Grafana 🔹 Tracing with OpenTelemetry & Jaeger 🔹 How the 3 pillars work together for real observability 💡 Observability isn’t about data — it’s about insight before impact. 🧩 Secret note: Even ChatGPT uses a layered trace + metrics system to monitor latency across its model calls in real time. 💬 Ever fixed a bug just by reading logs at 3 AM? 😅 #SystemDesign #Python #FastAPI #Prometheus #Grafana #OpenTelemetry #Observability #BackendDevelopment #DevOps

Radar 5 is built for usability—with Python flexibility and Generative AI power. It’s not just for data scientists. It’s for insurers ready to move faster and smarter. Read about Radar 5: https://ow.ly/CgXK50XkuGB Learn more about Radar with Python: https://ow.ly/OrlC50XkuGA #Radar5 #Python #InsurTech #Governance #WTWRadar

Learn how to use Python’s FastAPI framework to create an API that connects to and interacts with an OpenAI chatbot. This step-by-step guide walks through setup, integration, and testing—helping developers easily add AI-powered assistants to their applications. 👉 Read the full guide: https://bit.ly/4q6Fz4G #FastAPI #OpenAI #Python #APIDevelopment #AI #Chatbots

See a lot of posts about how Quant/Trading Algos: use Rust or C++? Neural Profit uses Python mainly because Standford's fork of Meta's Prophet; Neural Prophet uses Python (mainly due to PyTorch). When we forked it to add another 1000 commits on top of it...we cared about $ not Rust or C++ (plus we're low frequency trading, meaning performance for high frequency trading, isn't in our game plan). Will we refactor to Rust? Maybe one day...but I'll have someone else do that for me lol. P.S. see how much Shell is used? That's how much DevOps is needed on the platform and how significant that layer is in these Quant systems.

Tech With Tim: Build a Python AI Agent in 10 Minutes Fancy building a Python AI agent in 10 minutes flat? Tim’s got you covered: you’ll install dependencies, snag an OpenAI API key, import the right modules, set up your tools, spin up the LLM and agent, and run a quick test—all neatly timestamped. Grab the code on GitHub and follow along on YouTube. If you want extra hands-on support, his DevLaunch mentorship offers real projects and job-ready skills (no fluff). And don’t forget to try Notion for free to keep your dev life organized. #Python #AIAgents Watch on YouTube https://lnkd.in/gEkkD7fz

Radar 5 is built for usability—with Python flexibility and Generative AI power. It’s not just for data scientists. It’s for insurers ready to move faster and smarter. Read about Radar 5: https://ow.ly/ffP650Xkuym Learn more about Radar with Python: https://ow.ly/5SSx50XkunL #Radar5 #Python #InsurTech #Governance #WTWRadar