Start with the source.
A fixed financial snapshot with source references. Reported inputs stay separate from model assumptions.
Taiwan-based · Accounting background
Financial data,
made checkable.
I build Python workflows that connect financial inputs, calculations, and reports—so another person can follow the numbers and rerun the work.
One educational case. Clear assumptions. Inspectable code.
The approach
Easy to follow.
Possible to verify.
Make each step visible.
Cash-flow and net-debt bridges, followed by an explicit five-year growth path and valuation sensitivity.
Leave a way to check.
A readable memo, expected outputs, calculation tests, and commands for reproducing the Python case.
Selected work
Two ways to inspect my work.
01 · Public-data caseFrom statements
Try the interactive case From statements
to an analysis memo.
Explore Apple FY2025 cash calculations and change valuation assumptions in a browser. Compare the implementation with the Python source.
02 · Synthetic evaluationCheck financial
Inspect EvalForge Check financial
AI answers.
EvalForge combines deterministic checks, error categories, regression comparisons, and policy gates. Its published benchmark is synthetic and offline.
About
Accounting context.
Practical Python.
I’m Zixsa, based in Taiwan. My background brings accounting knowledge and systems experience to financial-data workflows.
This portfolio shows how I structure a problem, document assumptions, and make results reviewable. The examples are educational or synthetic; their scope is stated alongside the evidence.
- Junior Python automation and financial-data roles.
- Data validation, reconciliation support, and repeatable reporting projects.
- Finance-focused AI evaluation and QA tasks.
Remote employment
Accounting context.
Python contribution.
I’m interested in junior roles where financial understanding, careful data checks, and readable Python work are useful.
Roles I’m targeting
- Junior Python automation and financial-data analyst roles.
- Data validation and repeatable reporting support.
- Finance-focused AI evaluation and QA.
Evidence you can review
The financial case shows source-linked inputs, explicit assumptions, and tested calculations. EvalForge shows a synthetic evaluation workflow and its documented limits.
For a role discussion, share the responsibilities, eligible work locations, and expected working hours. I’m based in Taiwan.
Scoped freelance service
Financial CSV checks
& a summary report.
For an agreed financial CSV export, I can help check the data, document exceptions, and make the agreed calculations repeatable in Python.
Start with a small sample.
Share an anonymized CSV sample, a column description, the required output, and the totals or rules you want checked. Excel data can be supplied as a CSV export.
Make exceptions visible.
Agree checks for missing values, duplicate records, formats, and calculation or reconciliation differences. Exceptions are logged for review; ambiguous corrections need your decision.
Receive a repeatable result.
The agreed deliverables can include an output CSV, an exception log, a short summary report, and a Python script with run instructions.
Before work starts: we agree the file structure, checks, deliverables, acceptance criteria, fee, and deadline after reviewing the sample.
Scope: data preparation and agreed calculation checks. Accounting judgments, audit assurance, tax advice, and changes to live financial systems require a separate scope and appropriate expertise.
The linked report is an educational financial-analysis example showing traceability and reporting style.
A useful next conversation
Have a role or a reporting problem?
I’m interested in remote junior roles and scoped financial-data projects. For a project, start with sample inputs, the required output, acceptance checks, and a deadline.