Why did revenue decline in August?
- ✓ Establish revenue trend1 query · 0.4s
- ✓ Compare product categories1 query · 0.6s
- ✓ Compare customer segments2 queries · 1.1s
- ● Inspect cancellationsrunning
- ○ Inspect refunds
- ○ Validate strongest explanation
Everything you open a DB client for — browse tables, write SQL, export results — plus an AI analyst that investigates questions across your data and shows every query it ran.
“Why did revenue drop in August?”
SMB annual plans cancelled after the price change drove 71% of the drop.
DB clients were built to execute SQL. The hard part — deciding which query to run next — is still on you.
A stakeholder asks “why?” and you spend the afternoon writing throwaway joins.
You paste your schema into a chat window and hope the SQL it writes is right.
Everyone calculates “revenue” or “active customer” slightly differently.
One tool for SQL, another for charts, another for AI, a spreadsheet for the answer.
Ask in plain English. rowwise plans the investigation, runs read-only queries against your schema, and hands back a finding with the SQL, row counts and assumptions attached. Edit, cancel or re-run any step.
Why did revenue decline in August?
Revenue fell 12.4%. Cancelled SMB annual plans explain 71% of the drop.
SELECT segment, SUM(total) AS revenue FROM invoice JOIN customer USING (customer_id) WHERE status = 'PAID' -- from knowledge: Revenue GROUP BY segment;
Open tables in tabs, each on its own saved connection. Nothing auto-runs on restart.
Inline completion, formatting, query plans and saved queries. Writes are reviewed before they execute.
Turns questions into bounded read queries against your connected schema.
Save what “Revenue” means. The AI uses it every time. Export and share as YAML.
Pin saved queries to boards and revisit every past question in history.
CSV / JSON exports and saved snapshots you can re-run later for a fresh comparison.
Nine engines in one app, SQL and NoSQL side by side.
Add your own key and pick the model. Or keep everything on-device with a local model.
A desktop app, not a cloud dashboard. rowwise tells you plainly what an action will read, cost or change — before it happens.
Try everything free for 14 days. Then a one-time license unlocks every feature and database — no subscription, lifetime updates.
AI usage is billed by your own provider. rowwise never marks it up.
For day-to-day browsing, querying and exporting on the supported databases, yes. Some advanced features you may rely on — like SSH tunnels — aren’t in the beta yet, so many people keep their old client alongside while they try it.
Your data stays local. When you use Ask, schema details and your question go to the AI provider you configured. Result rows are not sent in follow-ups. Use Ollama to keep everything on-device.
No. AI-generated queries are bounded reads. Writes only happen from the SQL editor, after you review them.
Yes — OpenAI, Anthropic, Gemini, any OpenAI-compatible endpoint, or a local model through Ollama. The rest of the client works without one.
Every download starts a 14-day free trial with no account. After that, a one-time license unlocks everything — no subscription. AI usage is billed by your own provider, and rowwise shows the estimated cost before each investigation.
macOS and Windows.
Point rowwise at a database you already have. 14 days free, no account.