Managing your datascience projects is different from managing your typical IT projects. Here we provide tips on how managers can use AWS Sagemaker Feature store to save time and streamline the entire process of feature engineering across their datascience projects.

Todays episode, I will be talking to the founder and CEO of an AI Startup called MeghaAI (www.meghaai.com). Meghani has build a product where he claims to have automated the entire datascience pipeline for collecting industrial IOT data to building anomaly detection on it. This he claims helps many industries perform automated machine learning without the need of hiring datascientists. Lets listen to him talk about how he started his journey and how AWS has helped him build the product.

Todays episode we introduce you to machine learning models that have prediction errors, and these prediction errors are usually known as Bias and Variance. In machine learning, there will always be a deviation between the model predictions and actual predictions. The main aim of ML/data scientists is to reduce these errors in order to get more accurate results. In this episode we are going to discuss bias and variance, Bias-variance trade-off, Underfitting and Overfitting. Also, we would take a quick look on how AWS Sagemaker clarify helps us to understand data and model bias

Machine learning models are very different from code. When you deploy code you don't really need to monitor it on how it is delivering the results. However, ML models are different, we need to monitor their input data and measure them to a baseline. This is what we talk about in todays episode and talk on Services like AWS Sagemaker, Model Monitor, Model Drift and data collection. The process of Model Monitor is part of the MLOps lifecycle

If you are someone with zero programming skills and would still like to build ML models. Sagemaker Auto Pilot is the right tool for you. In this podcast me and deepti discuss a simple usecase of how to build a text classifier using Service Now dump. Listen on

MLOps is a set of practices for collaboration and communication between data scientists and operations professionals. Applying these practices increases the quality, simplifies the management process, and automates the deployment of Machine Learning and Deep Learning models in large-scale production environments. We talk about using AWS Sagemaker, Jenkins, Github and Stepfunction along with Sagemaker Pipelines

In computing, a pipeline, also known as a data pipeline, is a set of data processing elements connected in series, where the output of one element is the input of the next one. Here in AWS we talk about the data pipeline service and also talk about other services datascientists can use to build the end to end data processing pipelines for machine learning. The services we talk about are DynamoDB AWS S3 EC2 Data Pipelines Sagemaker

We all learn datascience in courses and get certifications but always in short of realtime projects. Here we talk how startups can build an end to end AI Webapp using AWS. We talk about an image classification example and talk about the end to end pipeline.Data ingestion, Data Labelling, Model Building, Model Inference and hosting a webapp. The services we talk about in this episode are AWS Amplify, AWS Lambda, API Gateway, Sagemaker, AutoML, Elastic Beanstalk and EC2

Having a visual flow chart of the entire piece of code is definitely a huge advantage while building a product. It enables a seamless way to debug at any point of time. AWS step functions step functions as a service, where in with a simple click of a button you could visually keep track of what is going on. Moreover, the console also highlights errors by which you could quickly pin point and trouble shoot them. In this episode, the discussion is to introduce this seamless approach to build Data science models and deploy them using AWS Step Functions.

Listen to Kali and Deepti talk about how SageMaker Data Wrangler can be used for data preparation and feature engineeringwalk you thorough use cases of data preparation workflows, including data selection, cleansing, exploration, and visualization. We talk about few examples in ver 300 built-in data transformations of SageMaker Data Wrangler so you can quickly normalize, transform, and combine features which saves a lot of time for the users without having to write code