Home

Six ways to break Excel wide open.

01

Say it.
It happens.

Charts, tables, formatting, formulas, sorting, filtering. Every Excel action you'd normally click through 4 menus to find — done in one sentence.

Every single operation is undoable with one click. You stay in control.

Create a bar chart comparing Q1 revenue by region, sort it descending, and highlight the top performer in bold

1Reading 4 columns, 120 rows from Sales Data sheet
2Grouped revenue by region, sorted descending
3Created clustered bar chart at F2:N18
4Applied bold + accent fill to EMEA (top performer)
02

Your spreadsheet has internet now.

Give it a list of names, companies, or anything. It goes to the web, finds what you need, and fills it in. LinkedIn URLs, funding data, headcount, emails — at scale.

You have

Company
Website
Industry
HQ
Stripe
Figma
Linear
Vercel

You get

Company
Website
Industry
HQ
Stripe
stripe.com
Fintech
San Francisco
Figma
figma.com
Design
San Francisco
Linear
linear.app
Dev Tools
San Francisco
Vercel
vercel.com
Cloud Infra
San Francisco

One prompt. Four columns filled. Works with hundreds of rows.

B2
=COASSISTANT.ASK("Classify "&A2&" as B2B or B2C")
B2B
C2
=COASSISTANT.WEB("Stock price of "&A2)
$218.47
D2
=COASSISTANT.ASK("Translate to Spanish: "&A2)
Informe trimestral
03

Two formulas.
Infinite power.

=COASSISTANT.ASK() puts AI in any cell. Classify, translate, summarize, extract — whatever you describe.

=COASSISTANT.WEB() pulls live data from the internet. Stock prices, company info, current events — grounded in real sources.

Drag to fill. Reference other cells. Nest with any Excel function. They're just formulas.

04

Drop a photo.
Get a spreadsheet.

Receipt? Screenshot? PDF report? Just attach it. CoAssistant reads the image, extracts every number, and puts it in the right cells.

Track expenses by photographing receipts. Digitize printed tables. Pull data from any visual source. Your phone camera becomes a data entry tool.

receipt_march_lunch.jpg

Attached image

“Add this to my expense tracker”
Date
Vendor
Category
Amount
Mar 6
La Boqueria
Meals
$42.80
05

When formulas aren't enough.

CoAssistant writes Python, executes it in a sandbox, and brings the results back into your sheet. Regression, clustering, fuzzy matching, custom visualizations — things Excel was never built to do.

generated by CoAssistant
import pandas as pd
import matplotlib.pyplot as plt
from sklearn.linear_model import LinearRegression

# Fit model on your sheet data
model = LinearRegression().fit(X, y)
r2 = model.score(X, y)

# Generate visualization
plt.scatter(X, y, color="#1a1a1a", alpha=0.6)
plt.plot(X, model.predict(X), color="#e63946")
plt.title(f"R² = {r2:.3f}")
Executed successfully · chart inserted at H2

Regression & forecasting

Fit models, predict trends, calculate confidence intervals

Fuzzy deduplication

Match near-duplicate entries that formulas can't catch

Custom visualizations

Heatmaps, cohort charts, anything matplotlib can render

06Coming soon

Excel talks
to everything.

Pull deals from Salesforce. Push status updates to Jira. Query your Postgres database. All through natural language, all from the sidebar. Your spreadsheet becomes the command center.

SalesforcePull open dealsREAD
PostgreSQLQuery last month's ordersREAD
HubSpotUpdate contact statusWRITE
JiraClose resolved ticketsWRITE
NotionSync project tasksREAD

This is what Excel
was supposed to feel like.

Free plan. No credit card. Cancel anytime.

CoAssistant