# Agentifying Legacy Systems

GLIK enables enterprises to layer intelligent automation on top of legacy infrastructure without requiring replatforming, middleware, or modern APIs. This section outlines how GLIK treats legacy systems as **agent surfaces**, making it possible to embed automation, policy logic, and AI-assisted decisions into tools that were never designed to support AI.

### Why Legacy Systems Resist Change

Most enterprise software was built before modern AI techniques were even imagined. These systems:

* Rely on brittle, vendor-specific APIs (or lack APIs altogether)
* Store data in formats that aren't machine-readable (e.g., PDFs, Excel exports)
* Depend on human approvals, manual routing, and tribal knowledge
* Are deeply embedded in financial, compliance, or operational workflows

This makes them **expensive to replace**, risky to modify, and difficult to integrate with newer technologies — including AI.

### GLIK’s Architectural Approach

Instead of forcing modernization through invasive upgrades, **GLIK wraps around legacy systems** with agent-based intelligence. It turns non-digital inputs (files, forms, chat, emails) into structured data, applies logic or LLM reasoning, and routes outcomes without touching the source system’s underlying code.

GLIK enables:

* **File-based control surfaces** (OCR, semantic parsing, classification)
* **Scoped memory orchestration** when APIs are unavailable
* **Human-in-the-loop workflows** for approvals and exception handling
* **Agent delegation and escalation logic** layered over static systems

GLIK treats any data format — from PDFs and CSVs to screenshots and inline chat requests — as an **actionable signal surface**.

### Agent Surfaces (Even Without APIs)

GLIK workflows can run across:

* PDF documents from legacy ERPs
* Excel-based report exports
* Manual audit logs or policy documentation
* Chat-based or email-based human approvals
* Static procurement, finance, or HR portals

These are processed using blocks like:

* `Doc Extractor` + `Parameter Extractor`
* `LLM` + `Question Classifier`
* `Knowledge Retrieval` from policy stores
* `Agent` blocks to simulate plugin responses

### Memory-Based Control (When You Can’t Integrate)

When API integration is unavailable, GLIK uses **variable memory** and **policy blocks** to simulate stateful decision-making.

* Memory variables store extracted inputs (e.g., `expense_total = 842.00`)
* Thresholds or blacklists are loaded from `Knowledge Retrieval`
* Decisions are composed with `LLM` + `IF/ELSE`
* Outputs can be submitted to a human or routed to another tool

This lets you deploy intelligent workflows **without real-time system integration** — perfect for pilot programs, compliance overlays, or operational triage.

### Benefits for Enterprise Developers

| Challenge                | GLIK Solution                                       |
| ------------------------ | --------------------------------------------------- |
| No accessible APIs       | Use documents or UI inputs as control surfaces      |
| Legacy forms/workflows   | Parse static inputs, classify actions, log outcomes |
| Risk of replatforming    | Wrap systems with agents, don’t replace             |
| Need for traceability    | Use audit blocks + variable memory + templates      |
| High cost of integration | Avoid middleware — deploy agents on top             |


---

# Agent Instructions: Querying This Documentation

If you need additional information that is not directly available in this page, you can query the documentation dynamically by asking a question.

Perform an HTTP GET request on the current page URL with the `ask` query parameter:

```
GET https://docs.glik.ai/system-architecture/agentifying-legacy-systems.md?ask=<question>
```

The question should be specific, self-contained, and written in natural language.
The response will contain a direct answer to the question and relevant excerpts and sources from the documentation.

Use this mechanism when the answer is not explicitly present in the current page, you need clarification or additional context, or you want to retrieve related documentation sections.