I made a manual Excel tracker for Upwork proposal costs, and I'm interested in how other freelancers keep their records useful.
One distinction matters: the cost of a winning proposal is not the total spent applying for work.
For a hypothetical completed set of 10 proposals, suppose each uses 14 purchased Connects, none are returned, and exactly one leads to a contract. At USD 0.15 per Connect, the set uses USD 21 in Connects, while the winning proposal alone uses USD 2.10. This is an illustration, not a typical win rate or earnings forecast. It excludes boosts, membership, other fees, taxes and time.
That's why my approach is to log every proposal, keep its status separate, and record earnings and withdrawals separately from contracts.
Kryzz AI is a free, MIT-licensed Android app for people who want an AI assistant without another account or subscription. You paste your own OpenRouter key, it uses hosted models (nothing runs on the phone itself), and your chats, memory, files and keys stay on the device. There's no Kryzz server in the middle.
It just went past 200 APK downloads from GitHub releases, all from posting and writing about it, no ads. This week it also got its first signed release build, about 19 MB instead of the old 82 MB debug one.
The honest weak spot is the first five minutes: getting an OpenRouter key is a step a lot of people don't want to do. If you tried it, or decided not to, I'd really like to hear where it lost you.
I'm working on TaskPaisa, a platform that connects users with online tasks and helps businesses and researchers reach registered users for survey participation.
Over the past few weeks I've been building Launch Pilot, an AI tool designed to help founders validate SaaS ideas before investing months into development.
Instead of using one generic AI prompt, it runs multiple specialized evaluations that look at areas like product clarity, validation, launch readiness, and potential risks before combining them into a structured report.
The goal isn't to replace user research it's to help founders identify obvious gaps and blind spots earlier.
Hey Product Hunt I m Shaylynn, founder of Gueaux and GueauxStudio.
GueauxStudio wasn t supposed to be its own product.
I was building Gueaux, a donation platform where those in need run campaigns. Campaigns with a video do better, so I set out to add a simple way for people to record one. Then I watched what happened: people who could talk about their cause for an hour froze the second the camera turned on. Scripts made it worse. They sounded like they were reading a hostage note.
For the past few days I've been building SmartReach, a part of xCinder, my workspace for people who run several X accounts. SmartReach looks for posts from people who need what you offer and helps you reply. It's still under heavy development and rough in places, so I'd like to share what I learned and hear how you handle the same problem.
Hey everyone, I'm Chitranshu, the maker of Where's My Therapist. We're opening our beta and I'd love your help testing it.
What it is
An all-in-one mental health app built for India. Instead of juggling an AI chatbot, a therapy app, and a separate psychiatrist, you get it all in one place:
Built Cairn alone for months. No co-founder, no funding, no safety net. Just me and the problem.
First launch ever. First upvote just came in 44 minutes ago from @david_firouzbakht Thanks david! If you're a solo founder who knows what this feeling is I'd love your support today. Upvote Cairn here: Cairn
