I built Delightful Cells after helping a friend extract cities from a spreadsheet full of inconsistently formatted addresses.
The name come from my impression to Excel. I don't use it at work and never learnt it properly. It feels daunting to me. So I want to build something delightful for it. Use simple natural language to get something done fast without fumbling with the UI.
It applies one AI instruction across hundreds or thousands of rows, with batching, retries, partial-failure recovery, and rerunning only failed rows. Deterministic tasks like filtering and replacing run locally in the browser. No Excel add-in or subscription required.
With Codex, Claude for Excel, Copilot, and other spreadsheet agents improving quickly, I’m unsure whether a focused tool like this is still useful. I’d love your honest feedback: what works, what breaks, or why you wouldn’t use it.
sumtsui•41m ago
The name come from my impression to Excel. I don't use it at work and never learnt it properly. It feels daunting to me. So I want to build something delightful for it. Use simple natural language to get something done fast without fumbling with the UI.
It applies one AI instruction across hundreds or thousands of rows, with batching, retries, partial-failure recovery, and rerunning only failed rows. Deterministic tasks like filtering and replacing run locally in the browser. No Excel add-in or subscription required.
With Codex, Claude for Excel, Copilot, and other spreadsheet agents improving quickly, I’m unsure whether a focused tool like this is still useful. I’d love your honest feedback: what works, what breaks, or why you wouldn’t use it.