Browse applications built on Mistral AI technology. Explore PoC and MVP applications created by our community and discover innovative use cases for Mistral AI technology.
Lightweight local security tool that prevents API key leakage by monitoring file changes in real-time and blocking git commits containing exposed secrets.
ContractGuard is a deterministic safety layer for AI coding agents that detects API breaking changes, maps their blast radius, guides automated repairs, and independently verifies release readiness with reproducible evidence.
VoxSales is a low-latency AI voice agent for automated lead qualification. Integrates Telnyx WebRTC, AssemblyAI, Node.js, and a Web Audio browser cockpit to process speech under 300ms, with post-call CRM and BANT automation via n8n.
An autonomous options-trading agent that sells overpriced volatility on SPY, QQQ and IWM and documents every time it refuses to trade, with an LLM desk that can only shrink or veto.
OrbiTrade is a multi-agent autonomous options trading system that resolves LLM hallucinations. It continuously improves by decoupling numerical programming from the LLM and learning from past trades.
An autonomous options-trading agent that combines quantitative market signals with multi-model AI, deterministic rule enforcement, risk controls, broker reconciliation, and autonomous execution.
StoryForge AI is a visual co-pilot for authors that cures writer's block by generating branching plot paths, scene outlines, and automated plot-hole detection while preserving the writer's unique creative voice.
AI tool to predict financial markets and automatically generate Python trading scripts using open-source LLMs.
Local-first AMD Track 1 agent that solves tasks with deterministic tools first and uses Fireworks AI rescue routing only when needed to preserve accuracy while minimizing tokens.
Building Human-like AI companions with human-inspired cognition and self-learning capabilities that grow, adapt, and evolve through experience.
Dockerized async video-captioning agent: ffmpeg keyframes->dual Gemini vision models describe scenes->rotating pool (Groq Llama, Mistral, Gemini Flash) writes 4 caption styles.Queue-based concurrency,retry/backoff, free-tier limits,10-min runtime budget.
ARIA is a multi-agent AI system that synthesizes biomedical literature from PubMed and audits it for citation fraud, methodology drift, overstated claims, cross-paper contradictions, and reproducibility gaps, built on Qwen2.5-72B via vLLM on AMD MI300X.
VibeCart AI is an AI vibe stylist that interprets your mood and curates a complete, shoppable outfit. Powered by AMD Instinct MI300X with 192GB VRAM. AI Model: Gemma 4 via Fireworks AI. We don't search catalogs. We style your vibe.
ASTRA is a hybrid Windows transcription app that filters audio locally, supports offline Whisper, and connects online to a license-protected server that batches jobs, routes them across speech providers, and automatically falls back when one fails.
PARALLAX is a GenAI-native cyber-fraud investigation platform that analyzes suspicious Android APKs, detects banking-malware behavior, coordinates agent review, and produces audit-ready action packets for fraud investigation teams.
WASI is an AI-powered WhatsApp food ordering platform orchestrated by Band AI Cloud. It uses a multi-agent architecture and Band's real-time telemetry to process Roman-Urdu, dynamically mutate cart state, and loop in human receptionists for edge cases.
Home-battery fleets following one price signal synchronise and breach grid voltage. GridAI is a four-agent protocol on Band: a Coordinator desynchronises dispatch, cutting herding overvoltage 471 to 0, with a full audit trail.
When a cargo ship stalls at anchor, the demurrage clock starts. Freight Room catches it from live AIS, convenes the right agents through Band, and hands an ops manager one costed recommendation to approve, with a full audit trail behind it.
OmniScope is an autonomous, multi-agent RevOps deal desk that transforms messy discovery call audio into compliance-audited, priced technical proposals in under 3 minutes using a self-healing double-rejection loop orchestrated over Band.
Five autonomous AI agents collaborate on Band to audit smart contracts — detecting vulnerabilities, simulating exploits, generating fixes, and publishing an executive risk report. Web3 security that thinks like an attacker, before $2.2B more is stolen.
BandAudit is an Agent Release Board to review release packets using provider-diverse AI agents, record evidence and votes in a Band room, and produce traceable approval, hold, and compliance reports for enterprise AI deployments.
An AI-powered recruitment workspace that extracts candidate profiles from resumes, ranks them against jobs, then runs autonomous voice interviews with 6 collaborative AI agents to produce evidence-based hiring decisions with full audit trails.
An autonomous, multi-agent DevOps pipeline that instantly intercepts, analyzes, and proposes patches for software crashes. Built on the Band SDK, it orchestrates specialized AI agents.
I built Bunobagera AI to review GitHub repos, chat about fixes, and automatically open PRs for production-ready code. It also audits your GitHub profile with personalized summaries and actionable guides to help you grow your developer brand.
Autonomous market research agent utilizing Gemini 3.5 Flash and Streamlit. It executes a parallel ReAct loop using Bright Data's SERP & Web Unlocker APIs to search, scrape, and compile structured company dossiers from live web data.
CodeSage is autonomous AI-powered DevSecOps platform that analyzes repositories, correlates vulnerabilities with live web threat intelligence using Bright Data and blocks unsafe deployments in real time through intelligent multi-agent security reasoning.
A drug-discovery research agent that searches PubMed for a target or disease, extracts compound candidates and mechanisms from each paper, cross-references clinical trial status, and returns a structured briefing with inline citations.
Upload any strategic document. Five adversarial AI agents — CFO, Market, Legal, Competitor, Execution — simultaneously attack it and deliver a severity-scored vulnerability report in under 90 seconds. demo credentials [email protected] red-team
An autonomous multi-agent contract auditor that moves beyond copilots into a decision-making system. It compresses a 15-hour manual legal contract review into a 12-second map-reduce parallel processing pipeline.
Code Janitor is an AI-powered code review and refactoring platform that detects bugs, suggests optimizations, visualizes codebases, and provides live previews, built with assistance from IBM BOB, IBM’s AI-powered development assistant.