It started when I was 19 and ran my first company. I had no idea what I was doing with money, so I found an experienced financial advisor who sat me down and showed me how companies actually do budgeting. Where money goes, how much you're spending, what income to expect. Basic stuff, but nobody had explained it to me before. I just started applying the same approach to my personal finances and never stopped.

Once a week I sit down, drop my bank statements into an AI agent, and it parses everything — categorizes transactions, inserts them into the database, checks that balances match across accounts. If something doesn't add up, it asks me before fixing it.

I tried a bunch of apps over the years — CoinKeeper, ZenMoney, spreadsheets. They all worked to some degree. But the one thing I couldn't find anywhere was the ability to just hand an API key to an AI agent and let it read, write, and modify my financial data however it wants. Every app wants you to click through its UI and use its import flow. I wanted SELECT, INSERT, UPDATE, DELETE through a simple HTTP API that any LLM can call.

So I built it. During a vacation. Was supposed to rest, but I couldn't stop once I started. Migrated all my data from ZenMoney and never looked back.

The main view is a budget table. Past months show actual numbers — what I really spent and earned. The current month has both actuals and the plan side by side, so I can see how I'm tracking. Future months are pure plan — expected income, expected expenses, projected balances. Everything sums up across categories and months, so I can see at a glance whether I can afford a big purchase three months from now or if I need to cut something. It's basically how companies do financial planning, but for one person.

Under the hood it's Next.js 16 with TypeScript on top of Postgres 18. Row Level Security enforced at the database layer — not in the app code. API keys are SHA-256 hashed, plaintext is never stored. Exchange rates fetched daily from five sources (ECB, CBR, NBS, NBU, plus a USDT peg). Locally everything runs in Docker Compose — Postgres, migrations, web app, FX worker. For production there's a full AWS CDK stack with ECS Fargate, RDS, ALB + Cognito auth, and WAF. The whole thing is about 9 tables and a view — kept it flat on purpose so LLMs don't get confused.

If you're into security — please try to break it. My real financial data is in there. I can't do a formal bug bounty, I'm one person. But if you get my data, I'll buy you a pizza and a beer.

Happy to answer questions about the architecture. Feature requests welcome too.

Comments URL: https://news.ycombinator.com/item?id=47215825

Points: 3

# Comments: 0

Our proprietary algorithm delivers unparalleled precision in risk assessment using advanced machine learning models. Beta testing demonstrates 99.999% accuracy in predicting negative outcomes.

Always says "don't invest, this will end badly"

Comments URL: https://news.ycombinator.com/item?id=44320656

Points: 6

# Comments: 0

We believe that AI will soon handle implementing most of the boilerplate and UIs in the specialized business software. Just look at what lovable.dev does today! Soon products which make creating business software easier for developers will become obsolete, or transform into "libraries" to be used by AIs.

We are losing this race, so we go the second route — publish everything and go on building other products on top of it.

GitHub repo URL: https://github.com/ozma-io/ozma

Ready to use: • GitHub repo, Apache 2.0 • Railway one-click deploy • AI-friendly code structure

Thanks to Nikolay Amiantov, Irina Gorokhova, Maria Podobrazhnykh, Gogi Gudzhedzhiani, Andrey Markin, Konstantin Remizov, Renat Darybaev, Leo Sobolev and Liubov Karas :)

Comments URL: https://news.ycombinator.com/item?id=43354889

Points: 6

# Comments: 0

You can deploy it to Heroku with just one click.

The bot handles text, voice, and photo messages, converting them to text for processing via OpenAI's ChatGPT.

It also uses PostgreSQL for data storage and optionally integrates with Pinecone for long-term memory.

The repository includes Docker and Docker Compose support for easy deployment.

Feel free to check it out, and let me know your thoughts!

Comments URL: https://news.ycombinator.com/item?id=40547401

Points: 69

# Comments: 54

We plan to put both back and front under Apache-2.0. It seems like a good enough way to show that the user won't depend on us.

> Will you be offering migration to and from your SaaS service?

Yes, we plan to continue offering the cloud for those who don't want to deploy their own solution.

> As for building a community, definitely make a Discord and a forum where people can come together and get help publicly, figure out how to get a stackoverflow space for it as well.

We're thinking about Discord right now, even started one. That is an interesting idea about Stackoverflow, thank you we'll think about it!