At a glance
- New challengers use AI to reinvent spreadsheet workflows, not just replicate Excel’s formulas.
- AI-native spreadsheets use conversational prompts to automate entire processes from data to summary.
- Other tools add an AI layer to Excel, automating repetitive bulk data tasks.
Creating an alternative to Excel feels like a fool’s errand. For almost four decades, founders have marched out with “Excel killer” decks and a fresh user interface – and the big green X has destroyed most of them. Only Google Sheets has survived, and then mostly by reproducing Excel in a browser window.
The latest crop of challengers takes a different tack. They are not trying to out-Excel Excel on formulas. These tools are built from the ground up to use AI for reinventing spreadsheet-based workflows that operate across multiple tabs and sources.
Yes, Microsoft has put Copilot inside Excel. It can help you write formulas, explain results and speed up analysis. But you still have to design the workflow and wire it together.
The new rivals fall into two buckets. AI-native spreadsheets try to replace the workbook with a prompt-first way of cleaning, modelling and reporting. A second group of tools keep Excel as the main interface and augment it with AI layers, adding functions to automate repetitive work.
AI-native spreadsheets
AI-native spreadsheets do something Excel still only does in part or clunkily with Copilot. If Excel works on a grid of formulas, AI spreadsheets are driven by conversational prompts. Their underlying AI creates and can execute repeatable workflows on your behalf. It often tries to run whole processes that accountants normally stitch together across spreadsheet data, Power Query, pivot charts and a separate written summary.
Say you get a 600-line dump of bank transactions and invoice lines. In Excel, you would typically use Power Query to clean merchant names, build a mapping table to categorise spend, create pivots by month, then write the “what changed” commentary in Word. Copilot can help generate a formula or explain a number, and Power Query can automate the clean-up once you’ve designed it. But Excel still expects you to design the process and keep it working.
In an AI-native tool, you can start with a prompt like: “Group duplicate-looking merchants, map them to eight expense categories, flag possible duplicate payments, then produce a monthly summary and a short variance note for the three biggest movements.”
The AI proposes the transformations, creates the outputs and saves the recipe so you can rerun it next month. The trade-off is trust: you still need to check results, because the AI can be wrong and it will not always show its working.
Paradigm pitches itself as a spreadsheet-based search engine which provides answers in a grid. But it also has specific use cases for accountants. It pushes conversational prompts and agents that can fill out and transform large amounts of data, rather than helping one cell at a time.
Paradigm combines prompts with research skills. That lets you order it to do things like: “Calculate the loan-to-deposit ratio for each bank in this table from their latest reporting data”.
Rows has another spin on your data. You can use prompts to join tables, add lookup columns and run multi-table analysis. It can also pull live data from other systems; essentially it comes with integration built in. Then you can use the AI features to analyse, summarise and build dashboards from whatever sources you have connected it to.
Want to step it up? Quadratic aims at power users. It blends the grid with code tools like SQL and Python, so teams can work with larger datasets and repeatable queries. You can use the AI as an interpreter and to draft queries and analysis.
Bricks feels less like a spreadsheet replacement and more like a way to quickly generate charts, summaries and presentation visuals from a spreadsheet full of data.
There are plenty of tools which already do this as part of more complicated programs – such as Aider, for accountants delivering advisory. Bricks, however, is more about getting a result quickly and easily; throw in your data and get a decent infographic that you can show in your presentation. Just remember to check the numbers …
AI inside the spreadsheet
This category assumes Excel and Google Sheets remain your weapon of choice, then adds an AI layer to speed up the boring parts.
Numerous.ai plugs into Excel and Google Sheets and lets you run AI across ranges. You can use it for classifying transaction descriptions, standardising supplier names or extracting key fields from invoice text. Excel can do parts of this with Power Query and Copilot, but Numerous is built around bulk “apply AI to this range” jobs.
SheetAI is essentially similar to Paradigm (AI-powered search inside a spreadsheet) but rather than having to use a new interface, you just use it inside Google Sheets instead.
Why bother using this over Google’s Gemini? SheetAI is much more flexible.
It can run AI from OpenAI, Claude, xAI and OpenRouter. It can also pull in information relating to video, OCR, image and audio through Replicate, an API-for-AI tool.
The latter may not sound that useful to accountants – until you realise that you could extract data directly from scanned invoices, photos of receipts and PDF statements, then drop the results straight into a worksheet.
Excel remains the unconquered giant; some estimates put its user base above one billion. But while it’s that big, innovators will keep aiming for a slice of its market.
Undertake flexible and online learning options with IFA Direct. The programme provides a direct pathway to IFA membership and enhanced professional recognition.
