Haku

Haku is a simple planner for Indian markets: set a yearly budget → allocate across stocks → let AI generate buy bands → log trades as you go → track plan vs remaining budget.

Most apps let you monitor existing investments. Haku is for planning them.

---

Contents

• The Problem
• The Solution
• AI Features
• Security
• Docs

---

The Problem

Most apps let you monitor existing investments — but the planning happens outside the app, on ad-hoc spreadsheets, scattered notes, and reminders. This means:

• An Excel sheet for per-stock allocation budgets
• Notes across devices for buy price targets
• No single view of: what's the plan, how much is deployed, is now a good price to add?

The result is a tracker to track your trackers — too much friction to follow consistently year after year.

---

The Solution

Haku brings the entire annual investment workflow into one place:

1. Plan your year — set a total budget, add stocks, assign allocation percentages by category
2. Buy bands — live financials fetched from Screener.in and NSE; valuation zones computed per stock by category
3. Log transactions — manual entry, grouped by month with buy/sell totals
4. Track deployment — see exactly how much of each stock's budget is deployed vs remaining
5. Carryover — unspent budget from one FY carries into the next plan automatically

No brokerage integration. No auto-sync. Intentionally deliberate.

---

AI Features

Band generation (no AI key needed)

Financials are fetched deterministically — no AI involved:

• Stocks: EPS, PAT, ROCE, and Mcap parsed from Screener.in consolidated annual P&L
• Index ETFs: index level and trailing PE fetched from the NSE API
• Auto-tranches: up to 5 buy tranches per stock, distributed within the Buy zone (skewed toward lower prices), qty sized from remaining FY budget

Investability scoring (Gemini)

The investability scorecard uses Gemini 2.5 Flash to score a stock across 10 qualitative gates (moat, management, balance sheet, etc.). To use it, add your own Gemini API key in Settings (tap the profile icon). Free keys available at aistudio.google.com.

---

Security

• All data stored in Supabase with Row Level Security (RLS) — users can only access their own data
• Gemini API keys stored encrypted at rest, never returned to the client after save
• No third-party analytics or tracking

---

Docs

Document	Description
docs/walkthrough.md	New user walkthrough — setup, screens, and flow
docs/app-spec.md	App model — vocabulary, screens, investment math
docs/valuation-playbook.md	Valuation rules — band multiples, signals, tranches
docs/architecture.md	Technical details — schema, band calculator, AI flow, security, setup
docs/design.md	Design system — typography, colour tokens, component contracts

---

Built with

Vibe-coded with Claude Code. Stack: Next.js 16, TypeScript, Tailwind CSS, Supabase, Gemini 2.5 Flash.