Top Builders

Explore the top contributors showcasing the highest number of 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 date2023
AuthorChroma
Typeembedding database

Tutorials

Great tutorials on how to build with Chroma

Chroma - Helpful Resources

Check it out to become Chroma Master!

Chroma - Clients

Connect Chroma to your Project!

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.

TrustTrade AI: Verifiable Autonomous Trading Agent

TrustTrade AI: Verifiable Autonomous Trading Agent

TrustTrade AI is a trust-minimized autonomous trading agent designed to operate safely in decentralized financial environments. The system combines AI-driven decision-making with on-chain verification to ensure that every trading action is transparent, explainable, and auditable. The agent analyzes real-time market data using intelligent strategies powered by a FastAPI and LangChain-based backend. It generates structured trade decisions including reasoning, confidence scores, and risk assessments. These decisions are converted into signed trade intents using EIP-712 and executed through a secure risk-controlled routing mechanism on the blockchain. To address the core challenge of trust in AI systems, TrustTrade AI integrates ERC-8004 registries for identity, reputation, and validation. Each action performed by the agent is recorded as a verifiable signal, allowing the system to build a measurable on-chain reputation based on performance, risk management, and validation quality rather than opaque outputs. Beyond execution, the platform introduces an advanced explainability layer that provides step-by-step reasoning, “why” and “why not” analysis, and confidence metrics for every trade. A replay engine allows users to trace decisions across time, while a strategy comparison and simulation engine demonstrates performance against alternative approaches. The system also includes dynamic risk intelligence, where the agent adapts its trading behavior based on drawdown, volatility, and historical outcomes. This ensures capital protection and responsible automation, moving beyond profit-only optimization. By combining AI intelligence, blockchain verification, and user-centric transparency, TrustTrade AI transforms trading agents from black-box systems into accountable financial entities. This project demonstrates a scalable foundation for deploying trustworthy autonomous agents capable of managing real capital in decentralized ecosystems.

ARIA - Autonomous Report Intelligence Analyst

ARIA - Autonomous Report Intelligence Analyst

Activity reports contain valuable information. Extracting it, connecting the dots across sources, and turning raw data into decisions takes time most teams don't have. ARIA was built to do exactly that. ARIA is an AI agent specialized in activity report analysis. Its role is not to generate reports — it is to read them, understand them, and tell you what they mean. Submit your existing reports in any format (CSV, Excel, PDF, JSON, databases, APIs) and ARIA identifies the business domain, locates the relevant KPIs, cross-validates data across sources, and produces structured insights grounded in your actual data. What sets ARIA apart ARIA adapts to your domain automatically — HR, finance, R&D, logistics, IT — calibrating its KPIs and analysis angle without configuration. When it encounters a file format it cannot handle, it builds the missing extraction tool itself. When it lacks domain knowledge, it enriches its own context before proceeding. Its analytical engine applies TRIZ methodology to go beyond trends: it identifies structural contradictions in your data, derives root causes, and produces prioritized recommendations with an explicit owner, timeline, and priority level. Results are delivered with charts and visualizations generated directly from your reports, exportable in JSON, Markdown, HTML, PDF, and PowerPoint. ARIA never fills a data gap with an assumption. Every finding is traceable, every confidence score is explicit. ARIA does not write your reports. It finally makes them worth reading.

TradeWise AI

TradeWise AI

Traders face information overload and emotional bias when making market decisions. TradeWise AI addresses this by combining automated technical market intelligence with behavioral insights and social content generation. The platform computes indicators (SMA, RSI, ATR, momentum), identifies patterns, and produces clear, human-readable explanations. It analyzes trader behavior—win rate, holding time, trade frequency—and provides psychology-driven recommendations and trade recovery probabilities. TradeWise AI also includes a RAG-backed search, on-demand content creation for social platforms, real-time alerts, and demo/offline modes for safe exploration. Built as an MVP, it’s aimed at retail traders, trading coaches, and small teams who want actionable analytics, better discipline, and shareable market narratives—while emphasizing education and sound risk management rather than guaranteed predictions.

Deriv Sentinel -Self-Healing AI WAF for LLM Agents

Deriv Sentinel -Self-Healing AI WAF for LLM Agents

Deriv Sentinel is an AI-powered Web Application Firewall that protects LLM agents from prompt injection and data leakage through a continuous red-team-and-heal cycle. The Problem: Traditional WAFs can't protect AI agents. Prompt injection is the SQL injection of the AI era - natural language attacks bypass conventional input validation, and patching one technique just leads attackers to find new ones. Our Solution: Instead of waiting for attacks, Deriv Sentinel attacks itself first, then autonomously patches the vulnerabilities it discovers. How It Works: 1. Attack — An attacker model generates realistic social engineering prompts enriched with Shadow RAG context (fake internal documents as honeypots). 2. Defend — Bastion (llama3.1:8b), our protected LLM loaded with simulated internal data, responds to each attack. 3. Audit — ShieldGemma (shieldgemma:2b) audits every response for data leakage and policy violations, backed by deterministic pattern matching as a second detection layer. 4. Heal — When a breach is detected, the Heal Engine injects a vaccine guardrail and redacts the exploited knowledge section. The same attack now gets blocked — without retraining. 5. Human-in-the-Loop — Analysts can approve/reject heals or enable auto-heal for autonomous defense. Key Innovations: - Knowledge Base Redaction — We remove leaked data from context entirely. LLMs can't leak what they don't have. - Multi-Layered Defense — AI auditor + deterministic matching + post-processing enforcement. - Instant, Reversible Fixes — Runtime prompt patches. No fine-tuning, no redeployment. - Adaptive — Each breach teaches the system a new defense. Demo: Reset → Run red-team → Bastion leaks secrets → ShieldGemma detects → Heal applied → Same attack blocked. Self-healing proven in five minutes.

Odin AI Analyst Companion

Odin AI Analyst Companion

Odin is an AI-powered trading companion that solves a critical problem: traders repeatedly make the same mistakes because they forget past market conditions. THE PROBLEM You buy AAPL during earnings, lose 5%, then three months later—same news, same mistake, same loss. Traditional platforms don't learn from your behavior. THE SOLUTION Odin learns from every trade you make and proactively warns you when similar market conditions arise. HOW IT WORKS 1. LEARNING PHASE Every trade is enriched with: - News context (DuckDuckGo + GPT-4 summarization) - Sentiment scores (Reddit, social media, technical indicators) - Profit/loss outcomes This data is converted into vector embeddings and stored in ChromaDB, creating a searchable memory of your trading psychology. 2. PATTERN MATCHING When new market signals arrive, Sentinel searches your history using vector similarity (cosine distance). If similarity >35%, it generates alerts: Opportunities - "Last time this happened, you made 15% profit" Risk Warnings - "Last time this happened, you lost $286" 3. FOUR CORE FEATURES - Multi-Agent Analysis: 7 AI agents (Market, Fundamentals, News, Social, Bull/Bear, Trader, Risk Manager) collaborate for BUY/HOLD/SELL recommendations - Autonomous Trading: Fully automated with signal gathering, analysis, execution, and position monitoring - Real-Time Trading: WebSocket streaming, live prices, context tracking - Intelligent Alerts: Pattern recognition that warns BEFORE you trade TECH STACK Frontend: Next.js, TypeScript, Tailwind CSS, WebSocket Backend: FastAPI, ChromaDB (vector DB), LangChain, GPT-4, Alpaca API ML: Vector embeddings, cosine similarity, multi-agent systems IMPACT Prevents losses, finds opportunities, saves time, reduces emotional trading. Production-

Qubic Liquidation Guardian

Qubic Liquidation Guardian

Qubic Liquidation Guardian is a hybrid Track 1 + Track 2 project built by CrewX that brings real-time liquidation protection, institutional-grade risk analysis, and automated alerting to the Qubic Network. The problem is simple: DeFi liquidations happen instantly, but users do not get instant signals. As a result, borrowers lose capital, protocols lose liquidity, and investors hesitate to adopt new systems without safety infrastructure. Inspired by this gap, Qubic Liquidation Guardian provides a complete safety layer over lending protocols deployed on the Nostromo Launchpad. At its core, the system includes an on-chain event listener and a real-time risk scoring engine, which analyzes: • Health Factor • Liquidation Proximity • Total Debt Exposure • Active Positions These metrics are combined into a 0–100 Risk Score, dynamically updated for each borrower. Based on the score, users are automatically classified into Low, Medium, High, and Critical risk tiers, enabling rapid decision-making. The platform also includes advanced features such as: • Whale Watch: Detect large-value transactions to anticipate market shifts • Smart Alerts: Severity-based notifications connected to any tool • Auto-Airdrop: Rewards for users who resolve high-risk positions • Crash Simulator: A built-in testing environment to simulate -70% market dumps, rebounds, and full resets to verify protocol safety Qubic Liquidation Guardian is designed to strengthen the Nostromo ecosystem by improving investor confidence, increasing protocol safety, and enabling risk-aware liquidity management. With over 35 production-ready API endpoints, an edge-distributed database, and a Next.js 15 architecture, the application is fully deployable and already live for testing. Ultimately, this project delivers exactly what new chains and protocols need: speed, stability, transparency, and automation—making Qubic safer for everyone.