Edge Analytics

We specialize in solving hard problems with undefined solutions at the intersection of bits and atoms. #edgecomputing #machinelearning #datascience

[Edge Analytics @_EdgeAnalytics]

Here @_EdgeAnalytics, we specialize in data science, machine learning, and algorithm development both on the edge and in the cloud. We provide end-to-end support throughout a product’s lifecycle, from quick exploratory prototypes to production-level AI/ML algorithms. 1/4

[Edge Analytics @_EdgeAnalytics]

Thanks to Connor Davis for putting this informative series together 👏 We are going to miss the generative images of pipelines! 🎨 3/3

[Edge Analytics @_EdgeAnalytics]

The final blog in our #MLOps series is out! If you're wondering how to take your #ML model out of a @ProjectJupyter notebook and into the real world, this blog covers everything you need to know. 1/3

[Edge Analytics @_EdgeAnalytics]

Learn about the key questions to consider when figuring out how to serve your model to users and explore different deployment methods, including @awscloud #Lambda and #SageMaker, as well as #TensorFlow Lite for #edge devices. 2/3

https://t.co/cLY2SMXsh5

[Edge Analytics @_EdgeAnalytics]

When running thousands of #ML trials, organization can get messy. That's why model tracking is essential for making high-value insights quickly. In our latest blog, we share how we integrate our #MLOps pipeline with @wandb for model tracking. https://t.co/c8vjdmm6Ij

[Edge Analytics @_EdgeAnalytics]

We built a serverless, event-driven app for handwritten notes summarization using @awscloud #Lambda, #S3, #Textract, and @OpenAI #GPT4. Check out the #tutorial and follow along to build your own app! https://t.co/2RaVnakAI0

[Edge Analytics @_EdgeAnalytics]

In part 3 of our #MLOps series, we discuss a few of the tools we like to use for model development. Follow along as we design and train a CNN w/ #TensorFlow, optimize hyperparameters w/ #KerasTuner, and set up a remote training session with #Ray. #ML #AI
https://t.co/caLuGpD8dC

[Edge Analytics @_EdgeAnalytics]

In the latest blog in our #MLOps series, we dive into the importance of cleaning, standardizing, and appropriately formatting raw data for #ML. We discuss how building flexible, robust, and scalable #data processing pipelines can be a game-changer! #AI https://t.co/lOtp4QIi8g

[Edge Analytics @_EdgeAnalytics]

In the first blog of our #MLOps series, we'll be focusing on data storage. 💾🔍
Learn how we store raw data for model training and evaluation in an @awscloud #S3 bucket and access it using the #Boto3 SDK in #Python. #DataScience #MachineLearning #AI https://t.co/aOdU5zWi8f

[Edge Analytics @_EdgeAnalytics]

Excited to share our #MLOps blog series! 💻🔬 Learn how investing in MLOps infrastructure accelerates insights extraction, optimizes iteration time & increases trust in models. https://t.co/yNVbJ6Bm8J #DataScience #MachineLearning

[Edge Analytics @_EdgeAnalytics]

In our latest article, @ren_gibbons shares how he leveraged the Edge process of rapid iteration and early prototypes to successfully build an #IMU based #wheelie prediction #ML model in <3 days. Get ready for a thrilling ride! 🚴‍♂️ #DataScience #AI

https://t.co/MuWB5G5GCA

[Edge Analytics @_EdgeAnalytics]

In our latest blog, Edge team member @RunsData explores how these same #generative models can be leveraged in #digitalmedicine to generate better #health outcomes.

https://t.co/dYNiaqWNtm

#generativeAI #AI #machinelearning #datascience #digitalhealth #healthcare

[Edge Analytics @_EdgeAnalytics]

Emerging generative machine learning models like #DALLE-2 and #ChatGPT have been making headlines for their remarkable artistic and conversational abilities. But their potential applications extend far beyond just entertainment.

[Edge Analytics @_EdgeAnalytics]

Our latest blog explores how we use the new #GPT-Edit model from @OpenAI to improve grammar + formatting in #OCR models for handwritten text transcription. Say goodbye to illegible scribbles and hello to formatted notes and summaries with GPT-Edit!
https://t.co/fxEGKyNMOk

[Edge Analytics @_EdgeAnalytics]

At @_EdgeAnalytics, many team members prefer handwritten notes. However, transcribing them to typed text can be painful when it's time to share. We end up either manually transcribing (⏰) or using OCR-based apps with disappointing results ("that doesn't look like my note!").

[Edge Analytics @_EdgeAnalytics]

Check out our new website to learn more about how we can help your business stay ahead of the curve in the ever-changing world of technology.

Have a hard problem in mind? Reach out! We are looking for our next inspiring client partner.

[Edge Analytics @_EdgeAnalytics]

Exciting news! Edge just got a makeover with a brand new website and updated logo! 💻🎉 Our focus on finding data-driven solutions to difficult problems is stronger than ever, and we can't wait to share our innovative process with you.
https://t.co/W00tUDqh7X

[Edge Analytics @_EdgeAnalytics]

His results show that #semanticsegmentation of roads in a #data limited context can be improved using generative models.
P.S. it has an associated @GoogleColab notebook for you to explore these questions yourself!

[Edge Analytics @_EdgeAnalytics]

Check out Edge’s new #blog written by Alex Browne! In it, he asks whether @StabilityAI #stablediffusion can be used to generate high quality labeled #data to make better #ML models in #computervision applications. #generativeAI
https://t.co/xApQsc0OYo

[Edge Analytics @_EdgeAnalytics]

2022 has been a journey! Highlights from the year include helping innovative companies create inspiring products, continuing to build a team of amazing people, and focusing on educational initiatives. Thank you for being a part of our journey! https://t.co/suDX5ndECW

[Edge Analytics @_EdgeAnalytics]

Did you know that you can use @huggingface models in popular #ML frameworks #PyTorch, #TensorFlow, and #JAX? Edge team member @_CarmenLai gives a tutorial on using transformer models from Hugging Face in these three popular ML frameworks. #datascience
https://t.co/Ye3PO8HTVG