> ## Documentation Index
> Fetch the complete documentation index at: https://docs.bluegenai.com/llms.txt
> Use this file to discover all available pages before exploring further.

# Platform Inputs Guide

> Find out what to gather and how to prepare it before starting your BlueGenAI project.

BlueGenAI accepts four categories of input. You do not need all four — use what you have. The more relevant context you provide, the better the platform output. But a plain language description of what you need to build is a valid and powerful starting point on its own.

***

## How to decide what inputs to gather

Start by identifying which scenario below best describes your situation, then gather the inputs listed for that scenario.

<CardGroup cols={3}>
  <Card title="Responding to a solicitation" icon="file-contract">
    Gather the RFP, task order, SOW, SOR, or Tech Spec documents. If you have discovery notes, interview transcripts, or JAD session outputs, include those as well.

    **Your entry point:** [Phase 1 Path A — Prototype from a Solicitation](/building-web-apps/phase-1-path-a)
  </Card>

  <Card title="Modernizing a legacy system" icon="arrows-rotate">
    Gather source code, database schemas, existing documentation, screenshots of the current application, and any ATO or compliance artifacts.

    **Your entry point:** [Phase 2 Path B — Legacy System Modernization](/building-web-apps/phase-2-path-b)
  </Card>

  <Card title="Starting from an idea" icon="lightbulb">
    You do not need any files to get started. Describe what you need in plain language — that description becomes your first prompt. BlueGenAI generates a complete, modern, accessible application design from that description alone.

    **Your entry point:** [Phase 1 Path B — Stakeholder Alignment](/building-web-apps/phase-1-path-b)
  </Card>
</CardGroup>

***

## Preparing your inputs

### Contractual Requirements

RFP, RFI, RFQ, SOW, SOR, or Tech Spec documents.

1. Locate your solicitation or contract documents
2. Export as PDF or keep as Word files — both are supported
3. Include the full document — do not excerpt or summarize before uploading
4. Upload via the Files icon in the toolbar

***

### Legacy Application Artifacts

Source code, database schemas, configuration files, and interface specifications.

1. Create a ZIP archive of the entire codebase — this is the preferred format for full codebases
2. Include configuration files in the archive: `package.json`, `requirements.txt`, `docker-compose.yml`, `.env.example`
3. Include database DDL scripts and schema exports
4. Include API interface specifications for any connected systems
5. Do not clean or filter the code before uploading — messy code is handled; removing files loses context

**Supported legacy patterns:**

| Pattern          | Examples                                     |
| ---------------- | -------------------------------------------- |
| Mainframe        | COBOL, JCL, RPG, AS/400                      |
| Legacy web       | Classic ASP, VB6, VBScript                   |
| Legacy Java      | J2EE, Struts, older Spring                   |
| Legacy .NET      | Web Forms, WCF                               |
| Database         | SQL, PL/SQL, T-SQL stored procedures         |
| Modern languages | Python, JavaScript, Java, C#, and all others |

***

### Support Documentation

Process flows, SOPs, policy documents, compliance artifacts, and screenshots.

1. Gather process flow documents, standard operating procedures, and policy documents
2. Take screenshots of existing application screens if available
3. Include ATO/Compliance documentation if you have it
4. Include system manuals, operator guides, and user guides

***

### Natural Language

User stories, meeting notes, JAD transcripts, and plain language descriptions.

1. Compile user stories, meeting notes, and JAD session transcripts into a document
2. These can be uploaded as files or typed directly into prompts — both work
3. Even informal notes and rough descriptions are useful

***

## Supported file formats

| Category             | Formats                                                          |
| -------------------- | ---------------------------------------------------------------- |
| Documents            | PDF (.pdf), Word (.docx), PowerPoint (.pptx)                     |
| Source code          | Individual files (.py, .js, .java, .cs, etc.) or ZIP archives    |
| Images & screenshots | PNG (.png), JPEG (.jpg / .jpeg), GIF (.gif)                      |
| Archives             | ZIP (.zip) — CAB extracts and categorizes contents automatically |

<Warning>
  Image-only PDFs with no text layer produce limited results. If your PDF is a scanned document with no selectable text, the platform may not be able to extract useful content from it.
</Warning>

***

## How to upload inputs to the platform

1. Open your project
2. Click the **Files** icon in the toolbar
3. Click **Upload** and select your files
4. Your files are now available to reference in any prompt

<Tip>
  You can also attach files directly in the chat input for immediate use in a single prompt. Drag and drop a file into the chat or click the attachment icon next to the prompt field. See [Navigation & Interface Guide](/platform-overview/navigation-and-interface-guide) for the difference between chat attachments and toolbar uploads.
</Tip>