Notebooks and Dashboards

Analyze in notebooks.
Share as dashboards.

Use the Alkera agent to create rich analyses in SQL and Python notebooks, then share them as interactive dashboards your team can explore.

From query to dashboard, in one notebook

Mix and match SQL and Python.

SQL cells query a connected warehouse and return a Polars or pandas dataframe by name. Later Python cells use it just like any other Python variable.

ordersCell 2 · SQL

Snowflake orders

SELECT order_id, account_id, order_month, amount_usd
FROM analytics.marts.fct_orders
WHERE order_month >= '2026-01-01'

FilterAll columns 48,213 rows

order_idUtf8	account_idUtf8	order_monthDate	amount_usdDecimal(12, 2)
ord_90412	acc_1182	2026-09-01	18,400.00
ord_90413	acc_0457	2026-09-01	2,950.00
ord_90414	acc_2210	2026-09-01	41,200.00

top_ordersCell 3 · Python

top_orders = (
    orders.filter(pl.col("amount_usd") > 10_000)
    .sort("amount_usd", descending=True)
)

Join data across warehouses.

Query one warehouse in one cell and a different warehouse in the next, then join the two results with SQL right in the notebook.

ordersCell 2 · SQL

Snowflake orders

SELECT order_id, account_id, order_month, amount_usd
FROM analytics.marts.fct_orders
WHERE order_month >= '2026-01-01'

FilterAll columns 48,213 rows

order_idUtf8	account_idUtf8	order_monthDate	amount_usdDecimal(12, 2)
ord_90412	acc_1182	2026-09-01	18,400.00
ord_90413	acc_0457	2026-09-01	2,950.00
ord_90414	acc_2210	2026-09-01	41,200.00

accountsCell 7 · SQL

BigQuery accounts

SELECT account_id, segment, region
FROM crm.accounts

FilterAll columns 3,402 rows

account_idUtf8	segmentUtf8	regionUtf8
acc_1182	Enterprise	North America
acc_0457	SMB	EMEA
acc_2210	Enterprise	APAC

revenueCell 8 · SQL

Notebook DuckDB revenue

SELECT o.order_month AS month, a.segment, a.region,
       SUM(o.amount_usd) AS revenue
FROM orders o JOIN accounts a USING (account_id)
GROUP BY ALL ORDER BY month

FilterAll columns 81 rows

monthDate	segmentUtf8	regionUtf8	revenueDecimal(14, 2)
2026-09-01	SMB	North America	145,672
2026-09-01	SMB	EMEA	85,127
2026-09-01	SMB	APAC	47,859

Chart the results.

Have your agent chart any data using Alkera's declarative chart library or the plotting library of your choice.

trendCell 11 · Python

alkera.chart(revenue.filter(period.contains("month"))).line(
    x="yearmonth(month)", y="sum(revenue)", color="segment",
).title("Monthly revenue by segment").tooltip()

Keep every result current.

Cells know what they read and what reads them. Move an input or change a query and dependent cells automatically re-run.

periodCell 9 · Python

period = alkera.ui.range_slider(
    months(revenue), value=("Jan", "Sep"), label="Months",
)

MonthsJan – Sep 2026

trendCell 11 · Python

alkera.chart(revenue.filter(period.contains("month"))).line(
    x="yearmonth(month)", y="sum(revenue)", color="segment",
).title("Monthly revenue by segment").tooltip()

by_regionCell 12 · Python

alkera.chart(revenue.filter(is_q3)).bar(
    y="region", x="sum(revenue)", color="segment",
).title("Q3 revenue by region").tooltip()

Hide the cells and code your readers don't need. The markdown, inputs and charts that remain become an interactive dashboard, still powered by the live notebook.

periodCell 9 · Python

period = alkera.ui.range_slider(
    months(revenue), value=("Jan", "Sep"), label="Months",
)

MonthsJan – Sep 2026

kpisCell 10 · Python

in_period = revenue.filter(period.contains("month"))
alkera.hstack([
    alkera.callout(f"**{usd(in_period)}** revenue", kind="success"),
    alkera.callout(f"**{share(in_period, 'Enterprise'):.0%}** Enterprise"),
    alkera.callout(f"**{usd(monthly(in_period))}** per month"),
])

$15.8MRevenue9 months

58%Enterprise shareJan – Sep 2026

$1.8MAverage per monthAll segments

trendCell 11 · Python

alkera.chart(revenue.filter(period.contains("month"))).line(
    x="yearmonth(month)", y="sum(revenue)", color="segment",
).title("Monthly revenue by segment").tooltip()

Run on the compute you need.

Run notebook kernels on a dedicated CPU or GPU node when you're crunching larger data or training heavy models.

finetuneCell 4 · Python

model = AutoModelForCausalLM.from_pretrained(BASE).to("cuda")
trainer = Trainer(
    model,
    TrainingArguments("ckpt", num_train_epochs=3, bf16=True),
    train_dataset=tickets["train"],
    eval_dataset=tickets["eval"],
)
trainer.train()
alkera.chart(losses(trainer)).line(x="step", y="loss", color="split")

Your agent builds it with you.

Ask for an analysis in chat. The agent writes, runs and fixes cells in the same live notebook you're looking at, and shows you the results along the way.

Build a Q3 revenue dashboard from Snowflake orders and our BigQuery CRM accounts.

I’ll join orders to accounts by account ID, then chart revenue by month, segment and region.

Createdq3-revenue.alknb.py8 cells
Ranq3-revenue.alknb.py8 cells · finished

Showed trend chartGo to cell

Enterprise drives most of the growth. Switch the notebook to Dashboard mode to share it with the team.

Turn your data into answers.

Start for free Book a demo

Back to product overview Read the documentation

Analyze in notebooks.Share as dashboards.

From query to dashboard, in one notebook

Turn your data into answers.

Analyze in notebooks.
Share as dashboards.