AWS SageMaker AI technology page Top Builders

Explore the top contributors showcasing the highest number of AWS SageMaker AI technology page app submissions within our community.

AWS SageMaker

Amazon SageMaker is a fully-managed machine learning service that makes it easy for data scientists and developers to quickly and easily build, train, and deploy models into a production-ready hosted environment. It provides an integrated Jupyter notebook instance for convenient access to data sources for exploration and analysis, eliminating the need to manage servers. Additionally, it offers optimized common machine learning algorithms that are designed to function efficiently with large data sets in distributed environments. SageMaker also allows users to bring their own algorithms and frameworks and offers flexible distributed training that can be adapted to individual workflows. Models can be quickly deployed into a secure and scalable environment using the SageMaker Studio or the SageMaker console.

General
Relese date	November 29, 2017
Author	AWS
Type	Learning Service

AWS Amazon SageMaker Libraries

Discover the AWS Amazon SageMaker libraries and SDKs.

AWS Amazon SageMaker Build, train, and deploy machine learning (ML) models for any use case with fully managed infrastructure, tools, and workflows
SageMaker Tutorials Machine Learning Tutorials with Amazon SageMaker

Edit on Github

AWS SageMaker AI technology page Hackathon projects

Discover innovative solutions crafted with AWS SageMaker AI technology page, developed by our community members during our engaging hackathons.

Act4City - Il ponte fra cittadini e PA

Problem I cittadini spesso incontrano difficoltà nel reperire informazioni o interagire in modo efficiente con la pubblica amministrazione. Allo stesso tempo, i comuni faticano a raccogliere feedback utili dai cittadini, portando a risposte lente e opportunità mancate per migliorare i servizi pubblici. Solution Act4City introduce agenti AI per colmare questa lacuna. Per i cittadini, un assistente su WhatsApp risponde in tempo reale alle domande utilizzando i dati estratti dal sito web di Ravenna. Le conversazioni vengono archiviate per analisi, consentendo ai comuni di affrontare le problematiche ricorrenti. Per le amministrazioni pubbliche, un chatbot web permette ai comuni di monitorare tendenze, raccogliere feedback e promuovere il dialogo con i cittadini, trasformando le interazioni in spunti utili. Target Audience La soluzione è pensata per le amministrazioni locali che desiderano migliorare l’interazione con i cittadini e semplificare la comunicazione, così come per i cittadini alla ricerca di risposte rapide e precise alle loro domande. Unique Features/Benefits Automated AI Assistance: i cittadini ricevono risposte immediate via WhatsApp, migliorando accessibilità e soddisfazione.. Citizen Digital Twin: i comuni ottengono preziose informazioni sui bisogni e le preoccupazioni dei cittadini attraverso l'analisi delle tendenze. Enhanced Dialogue: la piattaforma favorisce una conversazione bidirezionale, migliorando l'erogazione dei servizi pubblici e la trasparenza. Scalable Model: applicabile a diversi comuni per standardizzare e modernizzare le interazioni pubbliche. Questi flussi evidenziano come gli agenti AI semplifichino le interazioni tra cittadini e comuni, promuovendo una comunicazione più intelligente ed efficiente.

Tachiwin Indigenous Languages Translator

Tachiwin is an initiative aimed at building and developing Artificial Intelligence tools for Mexico’s indigenous and endangered languages. The goal is to provide their speakers and communities with tools available in other dominant languages, thereby balancing opportunities for development, access, and the use of technology. Additionally, it seeks to promote and revalue these languages and the knowledge of indigenous peoples who, until now, have been left behind in cutting-edge information technologies, all while maintaining a socially and ethically grounded vision. Developments Tutunakú-Spanish-English Translation Large Language Model We are among the first initiatives globally to achieve a language model for an indigenous language novel to the base model. In this phase, we have developed a pretrained and fine-tuned Llama 3.1 8b model, which, to the best of our knowledge, represents the first LLM implemented for a Mexican indigenous language. The model was trained using a custom instruction dataset containing translations between Tutunakú, Spanish, and English. We are currently preparing additional datasets to include Nahuatl, which will be released soon. This model now powers the new beta version of the Tachiwin Android App. The app has evolved from being a simple bilingual dictionary to a fully functional offline/online translator, enabling communities with limited internet access to use the tool offline or toggle between modes as needed.

ML-powered Malay Messaging Maker

Business Problem: MNCs lose $120 billion in annual revenue opportunities due to insufficient localization efforts in their promotions. Ineffective localization can hinge brand trust and cost brand reputation, while decreasing sales and customer acquisition. The Malay-speaking market is a strategic gateway to the broader Muslim market for MNCs. However, entering this market presents significant challenges, including lengthy agency hiring processes and potential cultural and religious misunderstandings. On top of that, traditional A/B testing is a slow and inefficient process, often taking 3-6 months to see the performance due to organizational bureaucracy and insufficient ML infrastructure. Content selection is frequently arbitrary and subjective, further complicated by the need for Malay language translation. Goals: To help businesses reach 290 mil. Malay speakers and enter the market with ease - leveraging ML-powered personalized messaging Solutions: Our plug-in platform automates the entire cross-cultural marketing process. From website analysis to message delivery, it ensures culturally sensitive communication with Malay speakers, eliminating the need for manual translation and human intervention. Technical approaches: - Use TogetherAI to preprocessed the dataset downloaded from HuggingFace - Fine-tuned Llama 3.2 using the refined dataset via TogetherAI - Used LlamaIndex to template prompts and execute the model

Memoir AI

Memoir AI is an innovative photo album assistant designed to revolutionize how we interact with our digital memories. Leveraging advanced artificial intelligence, Memoir AI offers a seamless, conversational interface that allows users to effortlessly search through their extensive photo collections. Users can upload entire photo albums, which are then meticulously organized and indexed by the AI, making it easier than ever to find specific moments. The app's intuitive design supports bulk uploads, ensuring that users can quickly and efficiently manage their photo libraries. With features like AI-driven tagging, facial recognition, and natural language search capabilities, Memoir AI not only enhances the accessibility of your photos but also adds a layer of personalization by allowing users to name and categorize faces within their albums. Our platform aims to provide a user-friendly experience, complete with the ability to integrate future features like linking to Google Photos and Apple Photos for even broader access. Through Memoir AI, we envision a future where your precious memories are always at your fingertips, organized and accessible through simple, conversational queries.

ClusterUp

I want to create a Connectors marketplaces between vectorized database like MongoDB and some complicated data querying platforms like Redpanda or Confluent. So I first trained my data and used server-less encoder to do RAG, and then in order to create a better environment node deployment processes to make any event faster. Based on this, the front end is wanting to create the marketplace assistant out and it is gong to be web application where people can find the API created uniquely and go ahead and deploy their own stuff. This idea comes into the difficulties I was running into when building RAG and trying to train my own LLM to do vertical or horizontal training and searching to anti-deepfakes.

Autilent

Autilent is a smart solution that is capable of automatically detecting high risk situations during driving and then helps eliminate them. A special device powered by autilent's proprietary AI algorithms is installed inside the cabin of the vehicle. This device has two cameras, driver facing and road facing. The video feed is analysed in realtime by our in-car processor and the system then detects potential accident-causing actions e.g distractions, drowsiness, mobile phone usage etc. Realtime alerts are then generated to warn the drivers and alert the command and control centres in order to take preventative action. Additionally, we also offer data driven insights into every individual driver's behaviour and give hyper-personalised recommendations in order to bring a positive change in the driver behaviour and make them safer on the roads. We are on a mission to save lives and resources and look forward to your support in this. In the demo, please select the dates 1-15th March, and the driver: Saad Farooq, and in the vehicle section, KSA-006