DecisionVault AI

DecisionVault AI is a GenAI-powered decision memory layer for teams and AI agents. It turns messy workplace communication into structured, reusable decision records.

Instead of only summarizing meetings, DecisionVault AI extracts what was decided, why it was decided, who owns the next step, who approved it, what workflow it affects, and what evidence supports it.

Product Positioning

DecisionVault AI is designed as an out-of-the-box decision memory MVP:

messy workplace communication -> structured decision records -> searchable decision vault

It is not just a meeting summarizer. The product focuses on durable business context that teams and AI agents can reuse later.

Why This Matters Now

Modern teams make decisions across meetings, Slack-style threads, email, project notes, incident reviews, and planning documents. The decision itself often gets buried, while the reasoning, owner, approver, and dependencies become difficult to recover later.

DecisionVault AI addresses that gap by turning unstructured communication into reusable decision records with source evidence and confidence levels.

What It Does

• Extracts structured business decisions from meeting notes, Slack-style threads, emails, project notes, and CSV/text exports
• Captures decision rationale, owner, approver, affected workflow, dependencies, source evidence, confidence, and reusable context
• Saves extracted records into a local decision vault
• Lets users ask questions across current extraction results or saved decision history
• Prevents simple duplicate saves
• Exports current and saved records as CSV or Excel
• Flags ambiguous items for human review

What Makes The GenAI Focus Different

DecisionVault AI uses GenAI to interpret messy human communication and identify decisions that are not always written in a clean format. The focus is on extracting:

• the actual business choice
• the reason behind it
• owner and approver context
• dependencies and conditions
• source evidence
• confidence level
• reusable context for future teams or AI agents

This makes the app more decision-memory oriented than a generic summarizer.

Saved Decision Vault

The saved vault turns one-off extraction into reusable organizational memory. Users can search previous decisions, filter records by keyword, confidence, or project/workflow, and download filtered or full vault exports.

Tech Stack

• Python
• Streamlit
• Gemini 2.5 Flash through google-genai
• python-dotenv
• pandas
• openpyxl
• Local JSON storage

Setup

Create and activate a virtual environment:

python -m venv .venv
.\.venv\Scripts\Activate.ps1

Install dependencies:

python -m pip install -r requirements.txt

Create a .env file in the project folder:

GEMINI_API_KEY=your_api_key_here

For Streamlit deployment, you can also set GEMINI_API_KEY in Streamlit secrets. The local vault file can be changed with:

DECISION_VAULT_FILE=data/decision_vault.json

Run Locally

python -m streamlit run app.py

Demo Files

For a quick demo, upload these files together:

• meeting_notes.txt
• slack_thread.txt
• email_thread.txt

Then click Generate Decision Memory.

For a more realistic workplace-style demo, upload files from sample_data/:

• sample_data/real_meeting_notes_anonymized.txt
• sample_data/real_slack_thread_anonymized.txt
• sample_data/real_email_thread_anonymized.txt
• sample_data/incident_decisions_anonymized.csv

These files are anonymized examples that mimic real decision-heavy workplace communication.

Upload Safety

DecisionVault AI validates uploads before sending text to Gemini:

• accepts only .txt, .md, and .csv
• enforces per-file and combined upload size limits
• rejects empty files
• rejects binary-looking files
• requires UTF-8 readable text
• validates that .csv files contain usable CSV rows

This MVP does not include antivirus or malware scanning. A public or enterprise deployment should add an antivirus service such as ClamAV or a cloud file scanning service before processing uploaded files.

Local Storage

Saved decisions are stored in:

decision_vault.json

You can override this path with DECISION_VAULT_FILE in .env or Streamlit secrets.

This keeps the MVP simple and easy to inspect. A production or team version should move this storage to SQLite, Postgres, or another managed database.

Deployment

The fastest deployment path is Streamlit Community Cloud:

1. Connect this GitHub repository.
2. Set GEMINI_API_KEY as a Streamlit secret.
3. Deploy app.py.

Render, Railway, Azure, or other Python-friendly hosts can also run the app with:

python -m streamlit run app.py

MVP Scope

This version intentionally avoids real Slack, Jira, Gmail, or document management integrations. The goal is to validate the decision-memory workflow first:

messy workplace text -> structured decision records -> saved searchable vault

Current MVP capabilities:

• upload .txt, .md, and .csv files
• extract decision records with Gemini
• save records locally in decision_vault.json
• search current records and saved vault records
• prevent simple duplicate saves
• export CSV and Excel files
• review low-confidence or ambiguous items

Not production-ready yet:

• no authentication or team workspaces
• no production database
• no enterprise access controls
• no live Slack, Jira, Gmail, or document integrations
• no formal compliance or data retention controls

Privacy Note

Avoid uploading confidential, regulated, or sensitive workplace data unless your Gemini/API setup is approved for that use. See PRIVACY.md for an anonymization checklist and data safety notes.