BestSendRate

• 

• 

• 

• 

• 

• 

• 

• 

• 

• 

• 

Inspiration

Sending money internationally should be simple, but finding the best deal rarely is.

People often have to open several money-transfer apps, enter the same details repeatedly, compare changing exchange rates, check fees, and search for promotions. Even then, the provider advertising the “best rate” may not be the one that delivers the most money to the recipient.

I built BestSendRate to turn that process into one search.

The vision is to create a global remittance search engine that uses real provider data, including active promotions, to identify the highest actual recipient payout for any supported country-to-country transfer.

What it does

BestSendRate allows a user to select:

• The country they are sending from
• The destination country
• The amount they want to send

The platform then compares data from major money-transfer providers and ranks the available options based on the amount the recipient will actually receive.

Instead of comparing only advertised exchange rates, BestSendRate considers:

• Live provider rates
• Transfer fees
• Promotional rates
• New-customer offers
• Supported transfer corridors
• The final recipient amount

The result is simple: users can immediately see which provider gives them the best real outcome for that specific transfer.

How we built it

I built BestSendRate with ChatGPT Sol inside Codex, using AI throughout the entire product-development process.

ChatGPT helped me move from an idea to a working product by acting as a technical architect, pair programmer, debugger, researcher, and product-design partner.

The application was built with a modern TypeScript and Next.js architecture. Provider information is processed through a structured data layer that normalizes different currencies, countries, fees, promotions, and exchange-rate formats into one consistent comparison system.

The ranking engine calculates the expected recipient amount for each provider and places the strongest offer first. The platform also stores structured rate information so it can grow into a historical remittance-data platform over time.

Rather than asking ChatGPT to generate the entire product from one prompt, I worked iteratively: planning each feature, implementing it, testing the output, identifying problems, and using Codex to improve the system.

Challenges we ran into

The greatest challenge was not simply displaying exchange rates. It was making completely different provider data comparable.

Money-transfer platforms structure their information differently. Some charge visible fees, while others include their margin inside the exchange rate. Promotions may only apply to new customers, specific amounts, selected payment methods, or particular transfer corridors.

I therefore had to design a normalized system that could distinguish between standard rates and promotional offers while still calculating a clear final recipient amount.

Other challenges included:

• Matching countries with the correct currencies and transfer corridors
• Preventing outdated rates from appearing current
• Normalizing inconsistent provider API responses
• Designing an interface that presents detailed financial data without overwhelming users
• Creating an architecture that can support many more providers and countries

ChatGPT Sol was especially useful when debugging these edge cases and restructuring parts of the application as the product became more complex.

Accomplishments that we're proud of

I am proud that BestSendRate is more than a static comparison page. It is the beginning of a real global remittance-data system.

The project includes:

• A working country-to-country transfer search
• Promotion-aware provider comparisons
• Rankings based on the recipient’s final payout
• A reusable provider-adapter architecture
• Structured data for currencies, countries, providers, and transfer corridors
• A design that makes a complicated comparison understandable within seconds

Most importantly, I built a functional product from an ambitious idea with ChatGPT as my primary development partner.

What we learned

I learned that the most difficult part of a financial comparison product is not writing the interface, it is building trust in the underlying data.

A comparison is only useful when rates are current, fees are included, promotions are clearly labelled, and users understand why one provider is ranked above another.

I also learned how effective AI-assisted development becomes when it is treated as an iterative collaboration rather than a one-command code generator. ChatGPT accelerated implementation, but clear product thinking, testing, and human judgement were still essential.

What's next for BestSendRate

The next step is to expand BestSendRate across more countries, transfer corridors, and money-transfer providers.

Planned features include:

• Rate-change alerts
• Historical exchange-rate charts
• Saved transfer corridors
• Personalised promotion eligibility
• Provider reliability and transparency information
• A public remittance-rate API
• Weekly and monthly market reports

Long term, BestSendRate aims to become the largest independent database of remittance rates, fees, promotions, and provider performance, helping individuals, researchers, journalists, and financial platforms understand the true cost of sending money around the world.

Built With

• chatgpt
• codex
• css
• next.js
• node.js
• openai
• postgresql
• provider
• react
• supabase
• tailwind
• typescript
• vercel