DEV Community

Asmae
Asmae

Posted on

Asguard โ€“ Blockchain-Powered Security & Intelligence with Redis 8 AI ๐Ÿ›ก๏ธ๐Ÿค–

Redis AI Challenge: Beyond the Cache

This is a submission for the Redis AI Challenge: Beyond the Cache.

What I Built ๐Ÿ› ๏ธ

Asguard is a next-generation decentralized security & intelligence platform designed for IoT networks, sensitive data pipelines, and enterprise systems.

We aimed to solve three problems simultaneously:

  1. Real-time threat detection using AI inference โ€“ Detect anomalies, security breaches, and suspicious IoT device behavior within milliseconds.
  2. Immutable event logging with blockchain โ€“ Guarantee that every logged event is tamper-proof, verifiable, and auditable.
  3. Lightning-fast decision-making with Redis 8 โ€“ Use advanced Redis features beyond caching to ensure ultra-low latency responses.

Our vision is to provide a secure AI assistant that can monitor, detect, and react to events in milliseconds, while ensuring data authenticity and trust through blockchain technology.


Demo

Imagine this scenario:

A connected factory has hundreds of IoT sensors sending telemetry data every second.

  • Redis Streams capture the incoming data in real time.
  • RedisAI runs inference on the data using pre-trained anomaly detection models.
  • When a threat is detected, an alert is sent instantly via Redis Pub/Sub.
  • The event is stored immutably on the blockchain for verification.

How I Used Redis 8 ๐Ÿš€

Beyond caching, Redis 8 was at the heart of Asguardโ€™s performance:

  • Redis Streams โ†’ For high-speed ingestion of IoT device data.
  • RedisAI โ†’ For executing AI models directly inside Redis, eliminating external API calls and reducing latency.
  • RedisJSON โ†’ For structured, queryable storage of processed events.
  • RedisSearch โ†’ For fast, indexed search of security events.
  • Redis Pub/Sub โ†’ For broadcasting real-time alerts to connected dashboards and monitoring tools.

This combination allowed us to process thousands of events per second, detect anomalies in under 10ms, and store results in a blockchain ledger without bottlenecks.


Why Blockchain? ๐Ÿ”—

Security is not just about detecting threats โ€” itโ€™s about proving that events were recorded faithfully.

By using blockchain, every event Asguard detects is hashed, timestamped, and recorded on-chain, making it impossible to alter without detection.


Tech Stack โš™๏ธ

  • Backend: Node.js + Express
  • Database & AI Engine: Redis 8 with RedisAI, RedisJSON, Streams, Pub/Sub, RedisSearch
  • Blockchain Layer: Ethereum-compatible network for immutable storage
  • AI Models: Anomaly detection trained in Python, exported to ONNX, run inside RedisAI
  • Security: Role-based access control, JWT authentication

The Future of Asguard ๐Ÿ”ฎ

  • Integrate federated learning to continuously improve AI models without centralizing sensitive data.
  • Deploy on edge devices for on-site, low-latency protection.
  • Expand blockchain integration for multi-party event verification in supply chains and government networks.

Team Members

Single submission by Asmae.


Thanks for reading and thanks to Redis for the challenge! ๐Ÿš€

Top comments (2)

Collapse
 
anik_sikder_313 profile image
Anik Sikder

Super clean architecture. Redis as both the AI engine and event pipeline is genius. How do you handle blockchain transaction throughput at scale?

Some comments may only be visible to logged-in visitors. Sign in to view all comments.