Python · Streamlit · AI Project 02

US + SE Asia
M&A Screener

A cross-border M&A screening tool covering 38 companies across five markets — the US, Indonesia, Singapore, Malaysia, and Thailand — with live FX normalization and an AI-generated deal memo for any company that clears the filters.

View Live App ↗
Type
Python App · M&A Tool
Markets covered
US, Indonesia, Singapore, Malaysia, Thailand
Date
July 2026
38
Companies screened across 5 markets
5
Screening factors (EV/EBITDA, growth, margin, leverage, market cap)
Live FX
USD normalization across all SE Asian currencies
AI ✦
Claude API deal memo for every qualifying company

Why I Built This

Most screening tools stop at the US or treat Southeast Asia as an afterthought. I wanted something that covers both on equal footing — screening for companies with a financial profile common among acquisition candidates across five markets, with currency-adjusted comparisons so a $2B Indonesian company and a $2B US company are actually being compared in the same terms.

The tool started as a Google Colab prototype and was rebuilt into a modular three-file Streamlit app with FX normalization, CSV export, a Claude API integration for deal memo generation, and a public GitHub repo.

What It Does

The Tool in Action

M&A Screener screenshot 1 M&A Screener screenshot 2

The AKRA Deep-Dive

To validate the tool wasn't just a black box, I took one output — PT AKR Corporindo (AKRA.JK), an Indonesian energy distribution and logistics company — and built a full manual investment memo on top of it, sourced from the company's actual FY2025 annual report rather than relying on the AI-generated draft alone.

AKRA surfaced with 19% YoY revenue growth, a 7.4x EV/EBITDA multiple, and a net cash position. The process also exposed a real data-quality issue I caught and corrected: the data provider reported ~34x debt/equity against the actual ~1.3x from the annual report. The memo also flagged a practical deal constraint — AKR's ownership is concentrated with a controlling family stake, meaning any real transaction would require negotiated buy-in rather than an open-market approach.

Built With

Python Pandas Streamlit yFinance Claude API (Anthropic) Live FX Normalization CSV Export

What It Doesn't Do


Project Links

Access the live app and source code.

⌥
GitHub Repository
Source code · Python / Streamlit
View ↗
PDF
AKRA_Investment_Memo.pdf
PT AKR Corporindo — manual deep-dive memo
View ↗