One Gateway. Every AI.
Every SaaS.
NeuralHub is the single source of intelligence for your platform. Manage AI models, agents, knowledge bases, and automations from one dashboard — integrate any SaaS in minutes.
// Before: scattered OpenAI calls everywhere // const reply = await openai.chat.completions.create({...}) // After: NeuralHub handles everything import { NeuralClient } from '@codeswayam/neural'; const ai = new NeuralClient({ apiKey: process.env.NEURAL_KEY }); export const reply = await ai.chat({ agentId: 'auraflow-support-bot', message, userId, }); // Guardrails, cost tracking, fallbacks — all handled.
10+
LLM Providers
<50ms
Gateway Latency
100%
Encrypted Keys
∞
Connected SaaS Apps
Everything AI. One Platform.
Stop managing AI scattered across every SaaS. NeuralHub centralizes models, agents, knowledge, and automation into one command center.
Unified LLM Gateway
Route to OpenAI, Gemini, Claude, or Ollama through one API. Multi-key pools, fallback chains, and automatic rate-limit rotation.
AI Agent Builder
Create named agents with system prompts, guardrails, and tool access. Deploy auraflow-support-bot that only answers brand questions.
RAG Knowledge Base
Upload PDFs, docs, and URLs. Agents get context via pgvector hybrid search with citations back to source documents.
Workflow Automation
Chain LLM calls, conditions, and API calls into multi-step pipelines. Runs async via BullMQ — no blocking your SaaS.
Guardrails Engine
Topic enforcer, PII masking, jailbreak detection, and output validation. Safety at the infrastructure level.
Usage Analytics
Tokens, cost, latency, and error rates per app and per agent. LangSmith-style full call traces for debugging.
API Key Management
Issue nhub_live_xxx keys per app. Scoped permissions, rate limits, IP allowlisting, and one-click rotation.
@codeswayam/neural SDK
Drop-in NPM package. Add AI to any SaaS in 3 lines. Same pattern as @codeswayam/auth — instant integration.
Multi-key pool per provider
Add multiple API keys per provider. NeuralHub round-robins across them — if one hits rate limit, traffic auto-routes to the next.
Live in four steps
From zero to production AI in minutes, not days.
Register Your App
Add your SaaS to NeuralHub and get a nhub_live_xxx API key in seconds.
Configure Agents
Build named agents with system prompts, guardrails, and knowledge bases in the dashboard.
Install the SDK
npm install @codeswayam/neural and call ai.chat({ agentId, message }) anywhere.
Monitor Everything
Track usage, cost, latency, and guardrail violations from the analytics dashboard.
Build agents that stay on-brand
The guardrails engine uses LLM-judged topic classification to enforce scope. Your AuraFlow support bot will never answer competitor questions or go off-topic.
- System prompt vault with versioning
- Topic whitelist & blacklist
- PII masking before LLM call
- Output JSON schema validation
- A/B test prompt versions with traffic split
auraflow-support-bot
gemini-1.5-pro · v2.1
System Prompt
You are AuraFlow's support assistant. Answer ONLY questions about AuraFlow features, pricing, and usage. Politely decline all other topics.
1,284
Requests today
420ms
Avg latency
17
Guardrail blocks
Guardrails
Simple, transparent pricing
Start free, scale as you grow. All plans include the full SDK and codeswayam-auth integration.
Free
Perfect for exploration
- 1 registered app
- 1 AI agent
- 10K tokens/day
- 3 providers
- Basic analytics
Pro
For growing SaaS products
- 5 apps
- 20 agents
- 2M tokens/day
- All providers
- Knowledge base (5 KBs)
- Workflow builder
- Full analytics
Enterprise
Unlimited scale
- Unlimited apps & agents
- Unlimited tokens
- Dedicated key pools
- SLA guarantee
- Priority support
- Custom model hosting
Your AI infrastructure,
centralized in minutes
Connect your first SaaS app to NeuralHub and ship AI features 10x faster. No scattered API keys, no duplicated agent logic.