Live tutorial · real results

From raw data to a
working automation, in 5 minutes.

Meet Maya, ops lead at a growing store. She has a database, a messy spreadsheet, and no time. Here's her actual first session — every answer below is real output from the product, querying the sample Acme Store dataset that ships with every install.

Connect Ask Upload Automate Share
1Connect · 60 seconds

Point it at your data.

Maya signs in — her workspace is provisioned on first login — and connects Postgres. If a credential is wrong, OEntity tells her exactly which field.

connect datasource
# testing a bad password first… POST /datasources/test → field: password code: auth_failed # fix it, retry… POST /datasources/test → ✓ connected · schema introspected
2Ask · 15 seconds

Ask it a question. In English.

No SQL, no dashboard to build. Maya just asks. OEntity reads the schema, writes the query, runs it, and answers from live rows.

chat · senso
Maya: How many customers do we have, what's our total MRR from active subscriptions, and which region has the most customers? senso · runs query_database … senso: Total customers: 24 Total MRR (active subscriptions): $438.00 Top region: APAC — 7 customers
→
Three real answers from the Acme Store data — no one wrote a line of SQL.
3Upload · 20 seconds

Bring the messy spreadsheet too.

Maya's quarterly targets live in a CSV. She drops it in and it becomes a queryable source — no import wizard, no schema mapping.

upload spreadsheet
POST /datasources/upload (targets.csv) → ✓ "targets" is now a live data source region · quarter · target_usd
4Automate · 30 seconds

Turn a sentence into an automation.

The moment it clicks. Maya describes what she wants in one sentence — OEntity compiles it into a real, saved workflow, then schedules it.

workflows · compile
Maya: "Every Monday, count the customers and email me a summary." compiling… workflow saved · 2 steps 1. count_customers (agent_step) └─▶ 2. email_summary (notify · email) # recurring schedule… POST /scheduled-tasks → ✓ next run: Mon 08:00 UTC
→
A description became a running graph. That report now writes and sends itself, weekly.
5Share · it's a team thing

Governed and multi-tenant from step one.

Maya's workspace is an organization with projects, per-session data grants, and a tamper-evident audit trail on every query. Her analyst joins with read-only access to one project; nothing leaks across tenants.

Straight talk — live vs coming

Everything above runs today. One thing on the way: chatting with your inbox ("triage my email"). OEntity connects mailboxes as a channel today; interactive email tools for the agent are the next build. We'd rather show you what's real than promise what isn't.

Your data already knows the answer.
Just ask it.

Connect a source, ask one question, automate the rest.

Start free