Click on "Deploy Server".
Wait a few minutes for the server to deploy. Once ready, it will show a "Started" state.
In the chat, type @ followed by the MCP server name and your instructions, e.g., "@RedisNexus check the health of the finspot tenant and suggest optimizations"
That's it! The server will respond to your query, and you can continue using it as needed.
Here is a step-by-step guide with screenshots.
The world's first AI-native Redis MCP server + SaaS platform. Nobody has built this before.
RedisNexus is an enterprise-grade platform that makes Redis operations intelligent by combining:
Redis MCP Server — 16 enterprise tools that let AI assistants (Claude, GPT, etc.) talk directly to Redis
Multi-Tenant SaaS Dashboard — Real-time monitoring for all your Redis instances across products
AI-Powered Intelligence — Cache strategy advisor, health scoring, anomaly detection
Kubernetes-Native — Production-ready with HA, auto-scaling, backups, Prometheus alerts
Related MCP server: Redis CRUD MCP Server
┌─────────────────────────────────────────────────────────┐
│ RedisNexus Platform │
├─────────────┬──────────────┬────────────┬───────────────┤
│ MCP Server │ Dashboard │ AI Engine │ Multi-Tenant │
│ (16 tools) │ (React SPA) │ (Advisor) │ (Isolation) │
├─────────────┴──────────────┴────────────┴───────────────┤
│ Redis 7.2 (Sentinel HA) │
│ Master → Replica 1 → Replica 2 │
├─────────────────────────────────────────────────────────┤
│ Kubernetes (K8s) │
│ HPA │ NetworkPolicy │ ServiceMonitor │ CronJob Backup │
└─────────────────────────────────────────────────────────┘Tool | Type | Description |
| Read | Auto-detect type, retrieve any key with metadata |
| Write | SET with TTL, NX/XX flags for distributed locks |
| Write | Atomic multi-field hash operations |
| Read | Selective hash field retrieval |
| Write | LPUSH/RPUSH for queue/stack patterns |
| Read | Range queries on lists |
| Write | Scored members for leaderboards |
| Read | Rank-based range queries |
| Write | Append to streams with MAXLEN |
| Read | Read stream entries by ID range |
| Write | Pub/Sub channel publishing |
Tool | Type | Description |
| Read | Production-safe SCAN (never KEYS) |
| Write | Batch 50+ commands atomically |
| Intel | AI-scored health analysis with recommendations |
| Intel | Keyspace pattern & memory profiling |
| Intel | AI advisor for caching architecture |
| Intel | Multi-tenant isolation & management |
| Read | Comprehensive server metrics |
| Write | Safe deletion with confirmation gate |
cd redis_nexus_mcp/
pip install -r requirements.txt
python redis_mcp_server.pypython redis_mcp_server.py --http --port=8000docker build -t santhira/redis-nexus-mcp .
docker run -p 8000:8000 \
-e REDIS_URL=redis://your-redis:6379 \
santhira/redis-nexus-mcp# Deploy entire stack
kubectl apply -f k8s/redis-nexus-full.yaml
# Verify
kubectl get pods -n redis-nexus
kubectl logs -f deployment/redis-nexus-mcp -n redis-nexusRedisNexus supports multiple products sharing the same Redis infrastructure with isolation:
tenant:finspot:* → Finspot Trading (NOREN)
tenant:linkedeye:* → LinkedEye ITSM
tenant:voicelead:* → VoiceLead AI
tenant:clinicvoice:* → ClinicVoice AI
tenant:hrassist:* → HRAssist AIEach tenant gets:
Key isolation via prefix-based namespacing
Usage tracking (API calls, memory, key count)
Resource limits (max keys, max memory, rate limiting)
Independent health monitoring
Use case: "trading API price cache for NOREN"
→ Recommends: Cache-aside with 1-5s TTL
→ Suggests: Sorted Sets for rankings, Streams for audit
→ Warns: Enable stampede prevention with SETNX locksHealth Score: 87/100
🟡 WARNING: Memory at 82% — plan capacity increase
🟡 WARNING: 3 slow queries detected (>10ms)
✅ Cache hit rate: 94.2% (healthy)
✅ Replication: 2 replicas connectedRedis 7.2 with Sentinel HA (auto-failover)
AOF + RDB persistence (dual durability)
Prometheus metrics + Grafana dashboard
PrometheusRule alerts (memory, latency, replication)
HPA auto-scaling (2-10 replicas)
NetworkPolicy (namespace isolation)
CronJob backups (every 6 hours)
TLS termination via Ingress
Rate limiting (100 req/min)
Non-root container execution
Health checks (readiness + liveness)
redis_nexus_mcp/
├── redis_mcp_server.py # MCP Server (16 enterprise tools)
├── requirements.txt # Python dependencies
├── Dockerfile # Production container
├── k8s/
│ └── redis-nexus-full.yaml # Complete K8s deployment
└── README.md # This file
RedisNexus_Dashboard.jsx # React SaaS Dashboard (artifact)Santhira (Rajkumar Madhu) — Founder & CTO Self-hosted, cost-effective, enterprise-grade.
RedisNexus: Because your Redis deserves intelligence, not just commands.
This server cannot be deployed
AI-Ready Data Platform that fills the missing layer between data management and AI execution.
Enterprise memory, search, and context for frontier AI. 38 tools for business intelligence.
Code intelligence platform for AI agents. 20 tools for architecture, security & impact analysis.
AI-native Day 0 modernization platform for PRDs, architecture, work orders, and code transformation.
