In this blog post, I break down LoRA, or Low-Rank Adaptation, a technique used to fine-tune large language models in a more efficient way.
JupyterLab is a popular IDE for Jupyter notebooks, code, and data. IIn this post, we share some of the best JupyterLab extensions that enhance JupyterLab's core functionality with data science and productivity in mind.
We are thrilled to announce the launch of Einblick Prompt AI for JupyterLab (ai-einblick-prompt in your extension manager). This revolutionary extension will transform the way you work with data.
A technical introduction to Einblick Prompt, a specialized code generation assistant built for data scientists. We underscore the significance of context in generating accurate code, highlighting the limitations of general AI tools for data science tasks. We share a few concrete examples illustrating the challenges in existing systems and our solution.
Here’s our 8-minute guide showing you how to move from Jupyter notebooks to Einblick canvases, and get much more productive.
Python notebooks only allow linear workflows, but the way we actually work and think is non-linear. Only a canvas gives you the grammar to express more complex relationships in a workflow and a visual way to lay out results.
With ChartGen AI you can create beautiful charts and graphs with as little as one sentence. In this blog post, we'll explore what ChartGen AI can do, how you can get started, and why AI-powered chart making is the next revolution in data visualization.
In the third installment of our series, I’ll be covering ChatGPT’s Code Interpreter. The brainchild of OpenAI, the company that brought generative AI to the forefront of everyone’s minds, the Code Interpreter is a plugin available within the ChatGPT chat interface for ChatGPT Plus users.
In the second installment of our series on the leading generative AI apps in the data space, we're diving into GitHub Copilot. This is the latest offering from everyone's favorite version control platform. Read on to learn more about GitHub's AI pair programmer.
How we create data visualizations is unnecessarily tedious, and using generative AI will allow us to create better results in less time.
We’re starting a series to cover the leading generative AI apps in the data space. In this article, we’ll be focusing on Jupyter AI, the OpenAI-powered extension available wherever the IPython kernel runs.
Regardless of your opinion of GenAI, it's something everyone should be learning about. So what is it? What are the use-cases, and what are the risks? We answer these questions and more in the following article.
There’s been a historically high barrier preventing the average insights consumer from accessing advanced data insights. But we are now in an exciting time, where the barrier to entry has been halved, due to generative AI.
We’ve compiled a list of companies and startups that are leveraging generative AI to create exciting products and services for teams across industries. Check out the companies below to save you and your team time.
Knowing how to code is great. Actually writing code is a slog. ChatGPT pairs harmoniously with Python as a code assistant, syntax know-it-all, and documentation reader. But ChatGPT is not a Jupyter notebook--it's not meant to be one.
Today we’re excited to officially launch Einblick Prompt, our context-aware AI agent. With Prompt in every canvas, our users can now generate entire data workflows with as little as one sentence.
Much of the joking about how frequently CEOs now say “AI” is anecdotal. We wanted to understand just how pervasive this trend is, and possibly understand a little bit more about who actually is talking about AI.
While there are plenty of articles discussing the hype, we want to talk about the hard problems facing engineers building production-ready software applications. Here, we share practical guidance from building Einblick Prompt, which is still in beta. We’re all building with large language models for the first time, and wanted to share our learnings ahead of launch in a few weeks.
In short, prompt engineering is tailoring the inputs fed into a large language model (LLM) in order to get a specific response back. Prompt engineering can generate more accurate or complete answers, lead to answers that are formatted in a particular way, or change the tone or style of the text.
In this post, we’ll attempt to answer the question, "What is an LLM?" in brief, as well as provide an overview of a few important large language models in recent history. Put simply, a large language model (LLMs) is a type of model that is trained on large datasets from various sources to learn the relationships between words, phrases, and sentences.
The Einblick team gathered in the one and only Salt Lake City, Utah to join in celebrating the 20th Anniversary of PyCon. Check out our recap of PyCon to learn about some of the cool innovations and conversations in the NLP, generative AI, and LLM space.
Although code has evolved greatly, since its inception, in a digital age, programming languages create barriers to accessing the power of programming. It is now time to radically rethink what code actually means. Does it have to be formal language, where we have to incant “public static void main string args” and the compiler moans about misplaced parentheses? I think not.
The space of LLMs is now a gold rush, meaning that innovation can be clouded. Our goal is to share semi-technical content that will help ground you factually in this fast-shifting hype cycle. This piece will cover four sections: transformers, GPT, how to work with the model, and how people are creating usable software from the model.
Today, Python is the dominant language in the fields of data science and machine learning, but this wasn't always the case. In this article, we will explore how Python became the language of data science, and discuss concrete examples of projects and innovations that were made possible thanks to Python.
In the language model world, the hottest topic is LangChain - a framework that you might be curious about if you haven't tried it yet. So, what exactly is LangChain?
By combining the latest advancements in generative AI with Einblick's visual canvas, we are excited to introduce our latest cutting-edge feature: Einblick Prompt. Einblick Prompt empowers users to construct data workflows using natural language prompts.
When Silicon Valley Bank was closed on Friday, March 10th, pretty much everyone we knew was shocked. Though, we are not finance experts, nor banking historians… but we are data people so we went searching for datasets and here's some findings about historical receivership.
Learn what my model predicts for the 2023 Oscars and how. By the way, bet small on Brendan Fraser and Michelle Yeoh–and it was super easy to call a win for Coda last year.
Data science is a creative design profession. Here’s why -- and what we as data scientists have to learn from designers and the visual arts. By embracing data science as an art form, we can take the best of both arts and science to push what is possible in the field.
Introducing Einblick’s revamped Python canvas. We’ve created for data scientists what Figma did for designers: an interactive canvas packed with the functionality of tools you love, while connecting everyone in the data science process.
We’re excited to introduce Einblick’s new AI Assistant: an in-line, context-aware wrapper around the OpenAI API–the technology behind ChatGPT. With just a few clicks, you can fix existing code, add comprehensive code comments, and even generate new code based on your own custom prompt.
Data scientists are often asked to be a “jack of all trades.” From domain expertise to probability and statistics, programming, machine learning, and more, you are always juggling tasks. But you also have to keep adapting. And yet another tool comes to the forefront of everyone’s minds: ChatGPT.
Data pipelines are the backbone of modern data management and analysis. Every second of every day we are creating tons of data, and we're also tracking all of them. The demand for data from end users grows every day. In this article, we'll provide an in-depth look at data pipelines, data pipeline management, including best practices and common challenges, so you can leverage your data pipeline’s full potential.
Glue code can be a major hindrance to data scientists, taking time away from core tasks and slowing down the entire process. Einblick offers a solution with its collaborative platform and visual data computing capabilities, helping data scientists work more efficiently and effectively. Try it for free on your next project.
With so much data, the question is always, how can we store it, and use it optimally, without wasting time, resources, and working hours. Data lakes and data warehouses are two widely used methods for storing and managing big data. Understanding the differences is crucial for organizations that want to make informed decisions about which type of repository is best for their specific needs and requirements.
Two of the most commonly used solutions are databases and data warehouses. While both of these solutions are used for storing and managing data, there are several key differences between them. In this article, we'll define databases and data warehouses, as well as how to choose between these two data storage solutions.
Automated machine learning, or AutoML, is a rapidly evolving subfield that can transform data scientists’ workflows and unlock new insights from data. In this article, we will explore the benefits and limitations of AutoML in data science, as well as its potential future developments and ethical considerations.
It's a brand new year, and it's an opportunity to reflect on what went well in the past 365 days and how I can do things better in the next 365. With mass-accessible tools like chatGPT, and increasingly solid data engineering, we are continuing to ride a wave of interest in data science and AI/ML, and that means more projects, more things to learn, and more efficiencies that we can find.
To succeed with supply chain analytics, companies need to build a strong data foundation, invest in the right technology and talent, and develop a clear business case for its adoption. Read on to learn more about the role of data in supply chain analytics, use cases, benefits, challenges, and trends.
In this article, we’ll talk about a few ways in which you can use AI and data science in the field of marketing, providing businesses with new ways to connect with customers and improve their marketing efforts.
Most of the work in the application of successful machine learning relies not just on the algorithms or the models themselves, but the logistics of it all. Many of these logistics fall under the umbrella of machine learning model management, or MLOps. In this article, we'll cover why MLOps is important, the AI project life cycle, challenges with MLOps, and techniques and tools to implement.
As data continues to proliferate in every part of our personal and working lives, it is increasingly important to understand how to bring data from disparate sources together. Data blending helps solve this precise issue. This article will explore the definition, importance, and process of data blending, as well as its benefits, challenges, and potential future developments.
As a part of informing your data and analytics governance strategy, it is important to ensure you are leveraging data lineage to the best of your ability. In this article, we will explore the concept of data lineage, its benefits and importance, and how it can be used in conjunction with data classification and data governance to improve data security and productivity.
In this article, we will provide a high-level overview of the four main kinds of analytics, with a focus on predictive and prescriptive analytics. Then we’ll go over a few common data analytics models that you may encounter as well as some industry-specific use-cases, so you feel empowered to start leveraging data science to propel your business forward.
Last week Einblick’s team returned from the AWS re:Invent conference, the company’s flagship conference for everything related to their cloud computing ecosystem. Here are our three main takeaways from the conference.
Einblick makes it easy for you to drop your data analyses into Notion. Simply take any Einblick dashboard and workspace and drop them into your Notion documents with an Embed block.
At Einblick, we are reimagining the data science workflow, and producing a next generation data science platform for the community. As a part of these larger pursuits, Einblick is thrilled to be entering the holiday season as one of the Python Software Foundation’s newest Partner-level sponsors.
In this post, we’ll be breaking down the what, why, and how of exploratory data analysis. We’ll start with a brief overview of what exploratory data analysis is, why it’s important, a high-level approach, and then we’ll dig into a specific example of EDA using a Goodreads dataset available on Kaggle. Throughout the example, we’ll cover some of the fundamental libraries for you to be successful as a data explorer.
Data cleaning as part of data preparation can involve many steps, tools, time, and resources. In this article, we’ll simplify the data cleaning process, and focus on how to clean data in Python using built-in packages and commands.
Part of the data wrangling process is to cleanse, aggregate, or otherwise manipulate data in preparation for analysis, visualization, or storage in a database. Read on to learn more about data wrangling.
Whether it's data preparation or going in-depth on the best steps to take to transform your data into something actionable, we have you covered. With that in mind let’s go over a comprehensive review of data cleaning.
Alteryx is a popular data science and analytics automation software program, but Alteryx can be a bit expensive. You may be looking for other alternatives, and want to understand the marketplace a bit better before committing to a solution.
Data exploration is the process of analyzing datasets to find patterns and relationships, and is sometimes more formally referred to as exploratory data analysis (EDA). Learn more about data exploration techniques that will help you build predictive models and craft compelling narratives.
Churn analytics is the process of measuring and understanding the rate at which customers quit the product, site, or service. Churn analytics is critical for getting a performance overview, identifying improvements and understanding which channels are driving the most value.
Data transformation tools help standardize data formatting, apply business logic, and otherwise play the transform role in ETL (extract, transform, load) or ELT (extract, load, transform). These tools are used to provide a more consistent, uniform execution of data transformations, regardless of data source.
Data profiling is the process of examining data from various sources and collecting statistics or summaries about the data. This process can help you check if you have the right kind of data for your problem, as well as ensure data quality.
In this post, I will highlight the core paradoxes that Gartner introduced to help the data and analytics community unleash innovation and transform uncertainty. By unifying seemingly disparate concepts, Gartner’s summit created opportunities for new perspectives on age-old problems.
We cover everything you need to know to get started with data preparation, so if you're already a data scientist or you are researching particular sub-areas in this field, this post is for you.
In a notebook, you can do a lot–from preprocessing data to EDA to tuning machine learning models–which is great! But, in notebooks, there’s a lot of upfront work that you, as data scientists, must do every time before, and as you start analyzing data and building models.
We conducted a survey about the top challenges facing data scientists and data professionals across industries. Remarkably few of the responses were about model accuracy, but much of it was around collaboration, process, and communication.
Data science notebooks are powerful, flexible tools that data scientists use every day. But they are code-heavy linear workflows which do not properly address data scientists' need for multi-stakeholder collaboration, reproducibility, fast iterative discovery, and operational work to deploy. We explore a few ways notebooks fail data scientists here.
Historically, Machine Learning algorithms were a bit painful to use, and required tedious human intervention in order to tune hyperparameters. Recent innovations in AutoML means that data scientists can now get better models in less time, by using new tools that support automatic exploration of how to assemble the best ML pipeline.
Low-code tools are revolutionizing businesses, enabling citizen developers to create new business applications that drive innovation. Now, the same thing is starting to happen for citizen data scientists.
As organizations made data analytics a strategic priority, demand for data analysis outputs exceeded supply of trained data scientists. To bridge the gap, no code workflow platforms (KNIME, Alteryx…) were developed to make advanced data science easier, and give access to wide audiences.
Move fast and break things — but still be data informed. Startups must tailor their data analytics practices to focus on on delivering strategic insights quickly. These are a few observations we’ve observed in our partnerships with startups, as Einblick helps lean organizations produce better analytics.
While code can accomplish everything, there is a set of repetitive operations where visual-based no code operators will help every data scientist. In that way, no code operators are just the next logical extension of importing libraries.
Why have advancements in Machine Learning (ML) imperfectly translated to better data driven decision making? How can business line stakeholders and data scientists bridge the gap between quality analysis and executed changes?
Here are some of the collaboration challenges in data science today, and a case study of how one of our clients implemented live co-working sessions to solve them.
As organizations empower democratized analytics, they must recognize how advanced tools like AutoML needs to be augmented with human intuition. Reducing the need to code has not invalidated the need for human-led explorations of data.
In data science, there are many different versions of correctness. Accuracy itself can be highly misleading: We don't want accurate nuclear launch detection and we don't want accurate self driving cars.
But it’s 2022 and it’s time to say goodbye to spreadsheets as the primary tool for data analysis. You should be able to work in a fast, collaborative space for business analysis, and harness innovations in AI/ML to quickly identify key drivers and even access predictive modeling.