You do not need to rewrite your application. OGuardAI is designed to slot into existing AI workflows with minimal changes. Most migrations require changing one line of code or adding a thin middleware layer.

Migration Paths

Current Setup	Migration Path	Effort
OpenAI / Anthropic API	Change `base_url` to OGuardAI proxy	1 line
LangChain agent	Add OGuardAI LangChain adapter	3 lines
Vercel AI SDK	Add OGuardAI Vercel AI adapter	3 lines
Python backend (FastAPI)	Add OGuardAI middleware	5 lines
Node.js backend (Express)	Add OGuardAI middleware	5 lines
Custom HTTP pipeline	Call transform before LLM, rehydrate after	10 lines
RAG pipeline	Use 4-step API (transform, index, search, rehydrate)	Replace ingestion step

Migration 1: OpenAI Proxy (Lowest Effort)

If your application calls the OpenAI or Anthropic API directly, point it at OGuardAI's proxy instead. No other code changes are needed.

Before:

from openai import OpenAI

client = OpenAI(api_key="sk-...")
response = client.chat.completions.create(
    model="gpt-4o",
    messages=[{"role": "user", "content": "Help customer Julia Schneider (julia@example.com)"}]
)
# PII sent directly to OpenAI

After:

from openai import OpenAI

client = OpenAI(
    api_key="sk-...",
    base_url="http://localhost:8081/v1"  # <-- only change
)
response = client.chat.completions.create(
    model="gpt-4o",
    messages=[{"role": "user", "content": "Help customer Julia Schneider (julia@example.com)"}]
)
# OpenAI sees: "Help customer `{{person:p_001:ad4f97591c16}}` (`{{email:e_001:1bcaef1a4aff}}`)"
# Response is automatically restored before reaching your code

Setup:

oguardai-proxy --target https://api.openai.com --policy default --port 8081

Streaming, tool calls, and all message roles are handled transparently.

Migration 2: Python Middleware (FastAPI)

Add OGuardAI as FastAPI middleware. Incoming text is tokenized; outgoing responses are restored. Route handler code does not change.

# Add transform/rehydrate calls around your LLM interaction:
from guardai_sdk import OGuardAIClient

guardai = OGuardAIClient(base_url="http://localhost:3000")

@app.post("/chat")
async def chat(message: str):
    # Transform: mask PII before sending to LLM
    result = guardai.transform(message)
    llm_reply = call_llm(result.safe_text)
    # Rehydrate: restore PII in the response
    restored = guardai.rehydrate(llm_reply, session_state=result.session_state)
    return {"reply": restored.restored_text}

Migration 3: Node.js Middleware (Express)

Same pattern for Express. Add the middleware; routes stay unchanged.

// Add transform/rehydrate calls around your LLM interaction:
import { OGuardAIClient } from "@oguardai/sdk";

const guardai = new OGuardAIClient({ baseUrl: "http://localhost:3000" });

app.post("/chat", async (req, res) => {
  // Transform: mask PII before sending to LLM
  const result = await guardai.transform({ input: req.body.message });
  const reply = await callLlm(result.safe_text);
  // Rehydrate: restore PII in the response
  const restored = await guardai.rehydrate({ output: reply, session_state: result.session_state! });
  res.json({ reply: restored.restored_text });
});

What Changes vs. What Stays the Same

Aspect	Changes	Stays the Same
LLM calls	Routed through OGuardAI (proxy or middleware)	Same API, same models, same prompts
Application code	1-5 lines added	Business logic, routes, database queries
LLM provider	Receives tokenized text instead of raw PII	Same provider, same billing, same models
Output quality	Preserved or improved (semantic tokens carry context)	Tone, style, language, format
Infrastructure	One new service (OGuardAI binary or container)	Everything else
Authentication	Add OGuardAI API key to config	Your existing auth unchanged

Rollback and Gradual Rollout

Shadow Mode (Staging Only)

Enable shadow mode in a non-production environment to evaluate OGuardAI without changing application behavior:

transform:
  shadow_mode: true

In shadow mode, OGuardAI runs the full pipeline but returns the original unmodified text alongside the protected version. Because the response carries raw PII, the server rejects shadow mode in production validation, so run it only in staging or against controlled, non-sensitive traffic while you compare outputs and verify detection accuracy.

Recommended Rollout Sequence

Pick one non-critical AI workflow and route it through OGuardAI in shadow mode in a staging environment.
Run shadow mode for 1-2 weeks. Review detection reports and compare outputs.
Set transform.shadow_mode: false to activate protection, then promote to production. Monitor for regressions.
Expand to additional workflows with per-workflow policies.
Add the NER sidecar if you need person/company/location detection beyond regex.

Emergency Rollback

Proxy mode: Change base_url back to the LLM provider URL. One line.
Middleware mode: Remove the middleware line. Redeploy.
No data migration needed. OGuardAI stores no persistent state. Removing it leaves your system exactly as it was.

Migration Guide