Your data, reasoned through by a team of AI analysts.
Four specialist agents — the Analyst, the Strategist, the Architect, and the Insights Engine — collaborate on every upload the way a consulting team would. Production-grade SQL behind every number. Two minutes from upload to a working dashboard.
Built for retail · FMCG · e-commerce · product analytics · No signup to start · 50 MB max
Revenue averaged ₹16.2L/day; weekday-weekend gap is the lever.
How the agents collaborate
Four agents. Three steps. Two minutes.
Your file passes through a pipeline of specialist AI agents that read, scope, design, and answer — collaborating end-to-end with no handoff loss between stages.
The Analyst reads your file
Excel, CSV, or Google Sheet. Multi-sheet workbooks, pivot artifacts, Devanagari column names, Indian comma grouping — all handled. The Analyst classifies every column and infers your business domain in seconds.
~ 10 seconds · The Analyst
StartThe Strategist scopes the dashboard
Two or three sharpening questions — what success looks like for you, which segments matter. The Strategist captures your operator intent so the dashboard answers the questions you actually have, not generic ones.
~ 30 seconds · The Strategist
StartThe Architect builds. The Insights Engine answers.
The Architect lays out KPIs, charts, and tables — every number backed by SQL you can inspect. The Insights Engine stays online for follow-ups ("why did Saturday spike?") with answers that query your live data, never guesses.
~ 90 seconds · Architect + Insights Engine
StartFor your business
Your dashboard, written in your business's language.
Not a generic BI tool with empty charts — a dashboard that already knows what matters in your line of work.
Retail
- Daily sales by store · top SKUs · basket size vs footfall
- Stock-outs vs deadstock — what's tying up working capital
- Hourly footfall pattern · staff schedule fit by store
“Spotted ₹4L of slow-moving stock at two stores. Moved it to the high-velocity branch in a week.”
FMCG
- Primary vs secondary sales — where the gap is leaking
- Scheme uplift by SKU · which discounts actually moved volume
- Outlet coverage — billed vs total in each territory
“Caught a ₹15L scheme that gave away margin without lifting volume. Killed it in the next cycle.”
E-commerce
- Channel margin after fees · own-site vs Amazon vs Flipkart
- Repeat-buyer cohort retention · CAC vs LTV by source
- Return-rate by SKU and category — which listings to fix
“Realised Amazon was 40% of revenue but 5% of margin after fees. Shifted spend to own-site, doubled it.”
Product analytics
- DAU · WAU · stickiness — and the cohort of users driving each
- Activation funnel · which step kills the most signups
- Feature adoption — who uses what, and who hasn't found it yet
“Found onboarding step 3 was killing 60% of signups. Cut it. Activation jumped 40% the next week.”
Production-grade trust
Every number traces to a query that actually ran.
Numbers on this dashboard come from deterministic code aggregating your real data — never from an LLM guessing. The AI picks WHICH columns to roll up. The actual SUM, AVG, GROUP BY runs in a separate compute layer, traceable end-to-end.
Every number traceable to SQL
Click any KPI or chart point — we show you the exact query that produced it. Paste-runnable in BigQuery, Snowflake, or Postgres so you can verify in your own warehouse.
Deterministic compute, not LLM guesses
Agents decide WHICH columns to aggregate. The actual SUM / AVG / GROUP BY runs in a separate compute layer — code, not a model. Numbers can't drift between answers.
Built by operators, not researchers
Designed by people who've shipped data products inside real businesses — so the agents know what a good month looks like in your line of work.
Pricing
Start small. Scale when you're ready.
Pro tier includes a 7-day free trial — no card needed. Cancel anytime.
Creator
For your first dashboardsDrop an Excel or Google Sheet. Get a working dashboard.
≈ ₹17 a day
Cancel anytime
- 5 dashboards per month
- Excel · CSV · Google Sheet uploads
- Auto-built KPIs, charts, and tables
- Export to PDF · share via link
- Email support
Analyst
For the always-on operatorAsk follow-up questions. Refresh Google Sheets in one click.
≈ ₹67 a day
7-day free trial · cancel anytime
Everything in Creator, plus:
- 10 dashboards per month
- Ask any follow-up question of your data
- One-click Google Sheets refresh
- Shareable dashboards (read-only links)
- Priority queue — 2× faster build
- Priority email support
Enterprise
For data-rich businessesA team of specialist AI analyst agents, built around your stack.
Volume + seat-based
- Team of specialist AI analyst agents
- Live connectors: Postgres · MySQL · Snowflake · BigQuery
- Salesforce · HubSpot · Zoho · Google Sheets sync
- SSO (SAML / OIDC) · audit logs · role-based access
- On-premise or private-cloud deployment
- Dedicated success manager + custom SLAs
FAQ
What founders ask before they upload.
Where is my data stored?+
Your file is parsed in your browser, then your rows + dashboard config + computed numbers are saved to our Postgres database (hosted in Bangalore). This lets you access your dashboards from any device, refresh from source, share with your team, and ask follow-up questions that run real SQL against your data. We don't train on your data and don't share it with any third party except the AI providers we route to (OpenAI, Anthropic — under their standard API terms, no training).
Can I delete my data?+
Yes. Delete a dashboard from your workspace and we cascade-delete everything tied to it — rows, config, computed digest, chat history. Delete your account and we wipe every record tied to your user_id. Email hello@dataeze.ai for a JSON export of everything we hold on you before you delete.
Which AI models actually run my queries?+
Mix of OpenAI (gpt-4o-mini for understanding, gpt-4o / gpt-5 for chat) and Anthropic (Claude Haiku for design, Sonnet / Opus for complex analytical questions). All under standard API terms — neither provider uses API traffic for training. See /privacy for the model-by-purpose breakdown.
What if my Excel is messy — pivot tables, footers, multiple sheets?+
The Analyst agent handles all of that: footer rows, mid-file Grand Total / Subtotal rows, multi-row headers, and pivoted layouts get unpivoted automatically. If something looked off, you'll see exactly what was skipped, on the dashboard.
Will it understand Hindi / Marathi / mixed-script columns?+
Yes — the agents read Devanagari column names and mixed-script values. Currency parses correctly whether your numbers say ₹, INR, Rs, or are formatted with Indian comma grouping (1,00,000).
Can I use the dashboard on my phone?+
Viewing dashboards and asking follow-up questions works fine on phone. Initial upload works best on a laptop, because parsing a 50 MB Excel file needs real RAM.
What happens if I cancel?+
You can export every dashboard to PDF or HTML at any time. After cancellation we keep your dashboards live for 30 more days so you can grab anything you missed, then archive them.
How fresh is the data after upload?+
What you upload is the snapshot. Re-upload (or click Refresh on Growth+ tiers) to pull a new snapshot whenever you want. Connector-based live sync is on the Enterprise tier.
Question not here?