talk-data.com

In this session, we’ll explore how to build and visualize scalable LangGraph workflows by integrating them with Reflex, creating interactive and polished proof-of-concept chat applications.

Learn how to move from simple Python scripts to robust, maintainable projects. Using a project as an example, I’ll share my process, along with practical tips for structuring code, organizing files, and writing maintainable applications.

EasyLens is an iOS app that uses AI and computer vision to simplify daily tasks like waste sorting, understanding German documents, and identifying tourist spots. Users simply point their camera and select a feature; the app does the rest using Core ML and OCR.

An AI-powered assistant that helps companies design sustainable products by leveraging Environmental Product Declarations (EPDs) for informed decision-making. The aim of the project is to simplify sustainability analysis and optimize product design for lower environmental impact.

This project aims to develop an intelligent system using computer vision to identify individual jaguars by their unique facial and body patterns. A Vision Transformer (ViT) and advanced self-attention models will be used for segmentation and classification, with fine-tuned embeddings to enhance accuracy. The system will aid zoologists in tracking jaguars, especially after natural disasters, and will be deployed as an API for practical use.

AI-based diagnostic tools for skin cancer have become increasingly popular due to their efficiency and low cost. However, a critical limitation of many existing models is their lack of diversity in training data—particularly the underrepresentation of darker skin tones. As a result, these models tend to perform poorly on non-white patients, increasing the risk of misdiagnosis or delayed diagnosis and contributing to significant health disparities. This project aims to address this issue by developing a diagnostic tool capable of accurately identifying malignant skin lesions across all skin tones

This project focuses on developing a speech-based screening tool for Alzheimer’s Disease (AD) and Mild Cognitive Impairment (MCI) using machine learning models trained on audio recordings from Spanish-speaking individuals. The goal is to create a non-invasive, cost-effective, and easily deployable diagnostic tool for early cognitive impairment detection.

On digital platforms (or in physical settings) where customers arrive repeatedly, platforms must decide which options to offer them, as this influences their subsequent choices as well as the outcomes for both user and platform. This project will 1) classify customers on a digital platform based on their latent motivation (think: whether they're here to learn vs earn, or which out of several options is the primary emotion driving them) using deep mixture models, 2) predict consumer actions by using deep learning to estimate parameters of the consumer choice model, and 3) use reinforcement learning to select customer offerings/recommendations in order to optimize a long-term outcome.

This project leverages graph neural networks (GNNs) to predict the antiviral potency of small molecules by analyzing their molecular structures and interactions. By learning complex patterns in chemical data, GNNs can identify promising drug candidates with high efficacy against viruses. The insights gained from this approach aim to accelerate antiviral drug discovery and development.

Send us a text Welcome to the cozy corner of the tech world where ones and zeros mingle with casual chit-chat. Datatopics Unplugged is your go-to spot for relaxed discussions around tech, news, data, and society. Dive into conversations that should flow as smoothly as your morning coffee (but don't), where industry insights meet laid-back banter. Whether you're a data aficionado or just someone curious about the digital age, pull up a chair, relax, and let's get into the heart of data, unplugged style! In this episode, host Murilo is joined by returning guest Paolo, Data Management Team Lead at dataroots, for a deep dive into the often-overlooked but rapidly evolving domain of unstructured data quality. Tune in for a field guide to navigating documents, images, and embeddings without losing your sanity. What we unpack: Data management basics: Metadata, ownership, and why Excel isn’t everything.Structured vs unstructured data: How the wild west of PDFs, images, and audio is redefining quality.Data quality challenges for LLMs: From apples and pears to rogue chatbots with “legally binding” hallucinations.Practical checks for document hygiene: Versioning, ownership, embedding similarity, and tagging strategies.Retrieval-Augmented Generation (RAG): When ChatGPT meets your HR policies and things get weird.Monitoring and governance: Building systems that flag rot before your chatbot gives out 2017 vacation rules.Tooling and gaps: Where open source is doing well—and where we’re still duct-taping workflows.Real-world inspirations: A look at how QuantumBlack (McKinsey) is tackling similar issues with their AI for DQ framework.

Send us a text Welcome to the cozy corner of the tech world where ones and zeros mingle with casual chit-chat. Datatopics Unplugged is your go-to spot for relaxed discussions around tech, news, data, and society. This week, we’re bringing you a special episode straight from RootsConf, our annual internal knowledge-sharing extravaganza! Hosts Murilo and Bart sit down with Tim and Ben, data strategy experts, for a lively chat about the state of generative AI as it transitions from a buzzword to a business tool. Highlights from this episode: Generative AI adoption: Are companies finally moving beyond pilot purgatory?The environmental cost of AI: Can emerging techniques reduce its heavy energy footprint?Bridging the knowledge gap: What’s missing for widespread AI adoption in organizations?Future trends: How generative AI might reshape personalization and business processes in 2025.Plus, we dive into the Gartner Hype Cycle and its relevance in understanding AI’s journey from innovation to disillusionment and beyond. Get ready to dive deep into AI’s evolving role and its impact on industries, sustainability, and society. Hit play and join the discussion!

Send us a text Welcome to the cozy corner of the tech world where ones and zeros mingle with casual chit-chat. Datatopics Unplugged is your go-to spot for relaxed discussions around tech, news, data, and society. Dive into conversations that should flow as smoothly as your morning coffee (but don't), where industry insights meet laid-back banter. Whether you're a data aficionado or just someone curious about the digital age, pull up a chair, relax, and let's get into the heart of data, unplugged style!

In this episode, we're thrilled to have special guest Mehdi Ouazza  diving into a plethora of hot tech topics: Mehdi Ouazza's Insights into his career, online community and working with DuckDB and MotherDuck.Demystifying DevRel: Definitions and distinctions in the realm of tech influence (dive deeper here).Terraform's Licensing Shift: Reactions to HashiCorp's recent changes and its new IBM collaboration, more details here.Github Copilot Workspace: Exploring the latest in AI-powered coding assistance, comparing with devin.ai and CodySnowflake's Arctic LLM: Discussing the latest enterprise AI capabilities and their real-world applications. Read more about Arctic - what it excels at, and how its performance was measuredMore legal kerfuffle in the GenAI realm: The ongoing legal debates around AI's use in creative industries, highlighted by a dispute over Drake’s use of late rapper Tupac’s AI-generated voice in diss track & the licensing deal between Financial Times and OpenAIFuture of Data Engineering: Examining the integration of LLMs into data engineering tools. Insights on prompt-based feature engineering and Databricks' English SDKAI in Music Creation: A little bonus with an AI generated song about Murilo, created with Suno

Send us a text Welcome to the cozy corner of the tech world where ones and zeros mingle with casual chit-chat. Datatopics Unplugged is your go-to spot for relaxed discussions around tech, news, data, and society.

Dive into conversations that should flow as smoothly as your morning coffee (but don't), where industry insights meet laid-back banter. Whether you're a data aficionado or just someone curious about the digital age, pull up a chair, relax, and let's get into the heart of data, unplugged style!

In this episode, we're joined by special guest Vitale Sparacello, an MLflow Ambassador, to delve into a myriad of topics shaping the current and future landscape of AI and software development: MLflow Deep Dive: Exploring the MLflow Ambassador program, MLflow's role in promoting MLOps practices, and its expansion to support generative AI technologies.Introducing Llama 3: Discussing Meta's newest language model, Llama 3, its capabilities, and the nuanced policy of its distribution, featured on platforms like Groq. Read more here.Emerging AI Tools: Evaluating Open-Parse for advanced text parsing and debating the longevity of PDF documents in the age of advanced AI technologies.OpenTofu vs. Terraform Drama: Analyzing the ongoing legal dispute between Terraform and OpenTofu, with discussions around code ethics and links to OpenTofu's LinkedIn, and their official blog response.The Future of AI Devices: Are smartphones the endgame, or is there a future for specialized AI wearables? Speculating on the evolution of AI devices, focusing on the Human AI Pin review, Rabbit R1 and Dot Computer Go check out the Youtube video after so you don't miss out Murilo in that suit he promised (with a duck tie of course).

Send us a text Welcome to the cozy corner of the tech world where ones and zeros mingle with casual chit-chat. Datatopics Unplugged is your go-to spot for relaxed discussions around tech, news, data, and society. Dive into conversations that should flow as smoothly as your morning coffee (but don't), where industry insights meet laid-back banter. Whether you're a data aficionado or just someone curious about the digital age, pull up a chair, relax, and let's get into the heart of data, unplugged style! In This Episode: OpenAI's Election Strategy: A deep dive into how OpenAI is approaching the 2024 worldwide elections. How OpenAI is approaching 2024 worldwide electionsTech Quirks: Exploring the idiosyncrasies of the Continuity Camera in Chrome. Is it a magic trick or just poor engineering?The Rise of Bluesky: Unpacking the significance of Bluesky's newly launched RSS feeds. Bluesky launches RSS feedsHot Takes on Data: Tackling the bold statement - Every data transform is technical debt. An exploration of what technical debt really means in the data world. Every data transform is technical debtJoin hosts Murilo and Bart as they navigate these intriguing topics, bringing their unique perspectives and a touch of humor to the table. Intro music courtesy of fesliyanstudios.com.

Send us a text Welcome to the cozy corner of the tech world where ones and zeros mingle with casual chit-chat. Datatopics Unplugged is your go-to spot for relaxed discussions around tech, news, data, and society.

Dive into conversations that should flow as smoothly as your morning coffee (but don't), where industry insights meet laid-back banter. Whether you're a data aficionado or just someone curious about the digital age, pull up a chair, relax, and let's get into the heart of data, unplugged style!

In today's episode Murilo & Bart discuss:

AI and Software Insights Introducing Gemini: Google's latest AI modelGoogle BlogFireship Dev TweetTechCrunch Article on Gemini DemoCommunication and Collaboration in Tech 6 tiny wording tweaks to level up your communication as a software engineerCareer CutlerMLOps and Model Development Navigating the chaos: why you don’t need another MLOps toolOpenLayer BlogChatGPT's performance on Julia vs. Python and R for LLM Code GenerationStochastic LifestyleEmerging Tech and Fun Finds JSONB in SQLiteSQLite ForumWizard Zines for a touch of geekinessWizard ZinesSports Illustrated's AI author sagaThe Verge ArticleMonaspace: A superfamily of fonts for codeMonaspaceHot Takes Paper: You Want My Password or a Dead Patient?Cohost ArticleIntro music courtesy of fesliyanstudios.com Check out the episode on YouTube.