File size: 2,425 Bytes
1a66f69 | 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 | ---
title: Finance Analyst (Concept)
emoji: 💰
colorFrom: green
colorTo: gray
sdk: static
pinned: false
license: other
---
# Finance Analyst — Document Review Platform (Working manager + concept analytics)
A browser app that pairs a **functional financial-document manager** with a
**concept analytics/AI layer** for monthly financial review.
> **Live demo:** https://finance-analyst-concept.vercel.app
>
> ⚠️ **Honest scope:** the document-management side is **genuinely functional**; the
> analytics/AI side is a **concept** (illustrative figures). See below.
## What actually works today
A real, self-contained client-side document manager:
- **Upload** financial documents (PDF / XLSX / CSV / images) by click or drag-drop.
- **Persist** them in the browser via **IndexedDB** — the actual file bytes are stored
and survive a refresh.
- **Organize** by period/month, with per-month counts and storage totals.
- **Download** or delete any stored file; clear a month or everything.
- A stats bar (active periods, total files, storage used, last upload) computed from
the real stored files.
## What is a concept (not yet wired up)
- The **KPIs, charts, "AI-generated findings", and extracted line items** are
illustrative sample data, not computed from your uploads.
- The **"AI analyst crew"** is a visual workflow demo; there is no LLM call yet, and
uploaded files are **stored but not yet parsed**.
## What would make the analytics real (roadmap)
- Parse uploads client-side — SheetJS (`xlsx`) for Excel/CSV and pdf.js (+ OCR for
scans) to extract tables/text.
- A normalization layer mapping parsed rows to accounts/categories.
- Compute the metrics (revenue, expenses, margin, cash flow, variances) from the
parsed data and drive the KPIs/charts/table from it.
- For genuine AI findings: a backend proxy calling an LLM (e.g. Anthropic Claude) on
the extracted data, so no key is exposed client-side.
- Real report export (PDF/XLSX) instead of the placeholder action.
## Tech
Single self-contained HTML file. Vanilla JS, Chart.js, IndexedDB. No build step.
Deploys as-is to any static host (Vercel, Netlify, GitHub Pages).
## Privacy
No API keys, no backend, no data leaves your browser — uploaded files live only in
your browser's IndexedDB.
## License
See [LICENSE](LICENSE). Provided for review and evaluation.
|