Chroma AI technology page Top Builders

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

Chroma

Chroma is building the database that learns. It is an open-source AI-native embedding database. Chroma makes it easy to build LLM apps by making knowledge, facts, and skills pluggable for LLMs. The fastest way to build Python or JavaScript LLM apps with memory

General
Relese date	2023
Author	Chroma
Type	embedding database

Tutorials

Great tutorials on how to build with Chroma

👉 Discover more Chroma Tutorials on lablab.ai

Chroma - Helpful Resources

Check it out to become Chroma Master!

Chroma Chroma website
Chroma Docs Documentation of Chroma
Chroma Repo Open-source Chroma repository

Chroma - Clients

Connect Chroma to your Project!

Chroma Python Python Client for Chroma
Chroma JavaScript High-Performance Chroma Client for JavaScript

Chroma - integration

Get started with Chroma!

Edit on Github

Chroma AI technology page Hackathon projects

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

VitalGuide

VitalGuide is an offline emergency assistance app designed to save lives during natural disasters or situations where connectivity is unavailable. The app addresses the critical need for reliable survival guidance and emergency first-aid instructions. It uses a Retrieval-Augmented Generation (RAG) system and the Gemma 2 model to provide quick and accurate answers by leveraging government datasets and trusted sources like PDFs and guides. This is important because FEMA's manual is large and unwieldy, and slower to leaf through than our LLM tool, when the seconds matter. Gemma 2 also has general knowledge that the FEMA manual doesn't have and can be asked follow-up questions if information is unclear. Specialized for individuals, communities, and first responders, it features multilingual support, ensuring accessibility for diverse users. Its unique offline functionality guarantees availability anytime, anywhere. VitalGuide empowers users with lifesaving knowledge, bridging the gap when conventional communication methods fail.

IoMarkets

Think of IoMarkets (www.iomarkets.io) as a Web3 crossover between J.P. Morgan and NASDAQ. Namely, IoMarkets Organisation (www.iomarkets.org) will be anchored by an AI-powered business intelligence platform and a regulated digital assets exchange with primary issuance & secondary market trading capabilities from the Wall Street technology Co, an institutional grade tokenisation engine, embedded payments, corporate finance advisory, investor relations and asset management arms. IoMarkets also plans to issue its own digital currency backed by investments in the best platform deals; launch private markets focused online news and internet TV outlets similar to Bloomberg. The grand vision/ultimate goal is to create a platform in which AI and blockchain blend together seamlessly. AI-enabled front end provides intelligent, adaptive user interactions, leveraging advance algorithms for analysis, generation and decision-making. This is integrated with a blockchain-enabled back end, ensuring data integrity and security through a decentralised ledger system.

EMO

Developed a groundbreaking AI-driven platform aimed at empowering children with Autism Spectrum Disorder (ASD) to enhance communication, social interaction, and emotional intelligence. The platform incorporates several features designed to create an engaging, supportive, and educational experience tailored to their unique needs. Central to the platform is an interactive chatbot powered by Llama 3.1, enabling dynamic, personalized text-based conversations. This chatbot helps children practice communication skills in a non-judgmental, adaptive environment, fostering confidence and improving social interactions over time. To promote cognitive growth and enjoyment, the platform includes games like chess and a memory game. These activities are designed to improve focus, strategic thinking, and memory retention while ensuring a fun and engaging experience. A core feature is the emotion-learning module, where children can participate in quizzes designed to teach emotional recognition. The quizzes present various images representing emotions, and children select the emotion they perceive. The platform provides detailed feedback on their answers, helping them understand their mistakes and learn effectively. Using the advanced capabilities of Llama 3.2 Vision, the platform also offers an image emotion analysis tool. Children can upload any image, and the AI generates an accurate description of the emotions displayed, helping them understand facial expressions, gestures, and social cues in real-world scenarios. This all-in-one solution combines the power of interactive AI and engaging activities to create a safe, inclusive, and enriching environment where children with ASD can thrive socially, cognitively, and emotionally. It empowers them with tools to improve their daily interactions and enhances their overall quality of life.

Found It

Aging parents often struggle with memory, making it hard to find everyday items, which can feel overwhelming. To address this, we propose an assistive AI-powered application leveraging Gemma models and video data to answer queries like "Where did I last see my pen?". The system processes video frames, generates embeddings, and stores them in ChromaDB alongside metadata for efficient retrieval. With a simple, user-friendly interface, users can locate items quickly and independently. Designed for elderly individuals and their families, this solution fosters autonomy, eases caregiving challenges, and offers peace of mind to loved ones living abroad.

Vox populi

Le proposte dei cittadini per migliorare i quartieri sono spesso trascurate o non ricevono una corretta prioritizzazione, rendendo difficile rispondere in modo efficace alle esigenze della comunità. La nostra soluzione consiste in un agente AI che analizza automaticamente le proposte, identifica i trend emergenti e suggerisce iniziative concrete. Grazie a Llama per l'analisi semantica e Groq per l'elaborazione rapida, l'assistente redige report, propone azioni e interagisce direttamente con i cittadini. Questo progetto è pensato per coinvolgere attivamente i residenti di Bologna, migliorando la comunicazione con l'amministrazione e promuovendo una partecipazione più trasparente e inclusiva per il benessere collettivo.

EDUCAT IA

EDUCAT IA es un app web educativa de herramientas que complementan el trabajo del docente y ayudan al estudiante, cuenta con: 1) CHATBOT educativo en el cual el estudiante podrá realizar todas las consultas referentes al tema tratado en la clase, podrá realizar un pequeño test e ira aprendiendo y brindando respuestas personalizadas a cada usuario, mantiene un historial de conversaciones. 2)Traductor Español a Ingles, ayudara al estudiante con el aprendizaje del idioma, el estudiante colocara el texto y la IA le proporcionara el detalle de la traducción palabra por palabra (encontrando la semántica adecuada) y además le brindará 02 opciones de traducción al texto (estamos trabajando para añadir voz y pueda el estudiante escuchar la pronunciación). 3)Chatea con tu PDF, herramienta con la que el estudiante podrá subir un archivo PDF la IA lo va a procesar, se cargara temporalmente en una carpeta y se almacenara en un BD Vectorial (CHROMA), hemos desarrollado un sistema RAG (utilizando Langchain) para que el estudiante converse con la IA sobre el contenido del documento y le brinde respuestas, resúmenes. En estos días de la hackathon Hemos realizado nuestras pruebas y el MVP en equipos WAWA Laptop (https://wawaperu.org/) los cuales tienen una placa SBC Raspberry Pi 4 de 8Gb de memoria RAM y se energizan con paneles solares portátiles, con los cuales hemos encontrado tiempos de respuesta aceptables(1/2minutos), estamos trabajando en un diseño para el case de la placa Raspberry Pi al cual le añadiremos una antena (para mayor alcance en el aula), con paneles solares portátiles y podrán ser colocados en escuelas rurales, periurbanos y urbanas. Para los modelos de IA tenemos como contenedor OLLAMA, ahí hemos desplegado los modelos LLAMA 3.1/3,2, para el RAG estamos utilizando "nomic-embed-text". Estamos testeando con los modelos LLAMA3.2-vision (11B) pero el tiempo de respuesta es muy elevado, seguimos trabajando por optimizar nuestra solución y añadirle mas herramientas.