The tool I built for myself first.

I used to send proposals to clients, and each proposal writing was costing me about 30 minutes (in a good day). The reason is each proposal needs understading the client's req, needs, and pain points. And then crafting a response that is not too long and not too short, but just right size and the most relevant content.

Feeding them to ChatGPT or Claude doesn't help, really. They keep missing some of the details that actually matter for each job but the worst part is (part that makes AI irrelevant for writing a proposal) they hand back a generic line, long and mostly unrelatable text that somewhat ignores prior experience and still needs rethinking and putting (just a bit less than handwritten version) effort into fixing.

I knew exactly what a strong pitch looked like, because I have sat on the other side of that decision - hundreds of times hiring for Meta and TikTok, and later running a side project (Practice Interview) helping engineers break into big tech, so I built something that paid attention to what the AI chatbots didn't.

Then I realized, what if I analyzed successful and failing proposals!
What if I could see the patterns and the mistakes?
I did so with the help of stats and AI analytics tools to understand winning patterns and losing patters. I called them green flags and red flags and started using it first to evaluate my own proposals then quickly to generate my own proposals. Now the RemoteSWE does all that for you too. RemoteSWE's agent is built using relevant proposal samples from real clients and projects.

Giorgi had been watching the same failure from a different angle, coaching freelancers through their first jobs on Upwork. Some became successful. Others struggled with proposal writing, even though he'd laid out the structure for them.

When they met and told each other what they do, it made sense to combine their knowledge and create a powerful tool that could empower freelancers to land more jobs.