Bar
Compare categories
The product
Everything you need, out of the box.
One AI analytics engine. Six capabilities that turn any data table in your product into an answer-and-chart surface — for every customer, on day one.
Capability 01
Natural-language queries.
Customers ask in plain English. tableArth.ai writes the SQL, runs it against your table, returns the answer. No SQL, no schema knowledge, no training data required.
- Schema-aware — understands your column types, joins, and filters.
- Multi-step questions, drill-downs, follow-ups in the same context.
- Works on tenanted data — the widget respects your auth.
tableArth.ai · Customer orders
AskDashboard
You
Which regions grew fastest YoY?
tA
APAC leads growth at
+48%, driven by new enterprise accounts in
Singapore and Tokyo. EMEA
follows at +22%, NA at +14%, LATAM at +9%.
Region · YoY growth%
APACEMEANALATAM
MEAFCISOC
Capability 02
The right visual, picked automatically.
Eight chart types, one engine. tableArth.ai chooses the right one based on the question and the shape of the answer — no chart config from your team or your customers.
Line
Trends over time
Area
Cumulative growth
Pie / Donut
Visualize proportions
Scatter
Reveal correlations
Stacked bar
Multi-dimension breakdown
Funnel
Process drop-off
Table
Detail & raw data
Capabilities 03 & 04
Built to feel live — and to never sit blank.
Streaming responses
Answers stream token-by-token as the model thinks, so your product feels like a live analyst on the table — never a spinner. Cancelable mid-stream, resumable on retry.
- → Token-by-token reveal under 200ms first-token latency target
- → Tracks SQL execution, chart generation, and narration in one stream
- → Cancellable; partial results never leak past the host
Suggested prompts
Every dataset comes with 15 smart analytical questions auto-generated by the engine — instant cold-start UX so users never face a blank box.
Top 10 by revenue
Churn risk this quarter
YoY growth by region
Outlier customers
Last 7 days summary
Capabilities 05 & 06
Dashboards on day one. Self-correcting SQL.
Auto-built dashboards
Every table you connect comes with a generated dashboard — KPIs, trend lines, and breakdowns, all chosen by the engine. Ship analytics in your product on day one without a BI tool.
Active users
1,284
+18%
Customers
42
+6 this week
Questions / 30d
9,402
↑ 22%
Retry intelligence
Generated SQL self-corrects up to three times before an error ever reaches the user — column name fixes, type coercions, join repair, and dialect adjustment.
attempt 1 · column "rev" not found
attempt 2 · rewrite → SELECT revenue
attempt 3 · ✓ executed in 240ms
FAQ
How the product works.
What kinds of questions can users ask?
Plain-English questions about the data in a table — totals, trends, breakdowns, comparisons, top-N, time ranges, and follow-ups. Users don't need to know SQL or the schema; they ask the way they'd ask a colleague, and tableArth.ai turns it into a query.
Does it write and run SQL automatically?
Yes. The engine writes the SQL from the natural-language question, runs it against your data, and returns the answer with a chart. No SQL knowledge is required from the end user.
What chart types does it support?
tableArth.ai auto-selects the right visual for each answer across bar, line, area, pie/donut, scatter, stacked bar, funnel, and plain table. You don't configure chart types — the engine picks the one that fits the question.
How fast are answers?
Answers stream token-by-token and typically resolve in under five seconds, so the panel never sits blank while a query runs.
What happens if the generated SQL is wrong?
The engine has SQL retry intelligence — it detects a failed or invalid query and self-corrects, retrying up to three times before falling back gracefully, so a single bad query doesn't surface an error to your user.
Do users have to build dashboards?
No. Every table gets an auto-built dashboard on day one, and there are suggested prompts (15 per dataset) so users have somewhere to start without configuring anything.
Ship it
Plug it into your product. Today.
Pick the deploy that fits — Widget, API, or Chrome Extension — and ship to one dataset first.