Friday
Powered by Fridaybot
Your question, the context, the code, and the published answer — together in one traceable workspace.
Private beta · Free during beta · No credit card · Local-first
Developers have the IDE — code, context, execution, and output in one place — but data teams never got the same for analytics. Friday is that workspace: ask a question, attach governed context, run the SQL, pin artifacts, and publish with the full trace still intact.
Source trace stays attached from first question to shared answer — every number traces back to the SQL and tables that produced it.
Stop re-explaining the same question across chat, docs, notebooks, spreadsheets, charts, and slides. Friday gives data professionals one workspace built for analytics, where context, code, artifacts, and source trace stay together.
Friday workspace
Which segment drove Q4 expansion?
revenue_events, customer_segments, NRR memory
SELECT segment, SUM(revenue)...
Dashboard and PPTX with source trace attached.
Friday first attaches the context needed to answer the question: governed tables, metadata, KPI rules, business memory, and allowed scope.
Why did revenue slow down this quarter?
revenue_events net revenue after refunds and credits
customer_segments quarter-start segment assignments
accounts account ownership and firmographics
subscriptions lifecycle stage and renewal timing
Ask in plain English while table mentions, project memory, generated SQL, Friday's reasoning, and pinned artifacts stay together.
Why did revenue slow down this quarter?
revenue_events net revenue after refunds and credits
customer_segments quarter-start segment assignments
accounts account ownership and firmographics
subscriptions lifecycle stage and renewal timing
SELECT segment,
SUM(CASE WHEN fiscal_qtd = 'previous' THEN net_revenue_usd END) AS previous_qtd,
SUM(CASE WHEN fiscal_qtd = 'current' THEN net_revenue_usd END) AS current_qtd
FROM revenue_events
JOIN customer_segments USING (account_id)
WHERE fiscal_day_index <= current_fiscal_day_index
GROUP BY segment;WITH segment_qtd AS (
SELECT fiscal_quarter, segment, SUM(net_revenue_usd) AS revenue
FROM revenue_events
JOIN customer_segments USING (account_id)
GROUP BY fiscal_quarter, segment
)
SELECT segment, current_change, avg_change, sigma_from_avg
FROM segment_qtd_signal;SELECT lifecycle_stage,
SUM(net_revenue_usd) AS current_qtd,
lifecycle_range_status AS normal_range
FROM revenue_events
JOIN subscriptions USING (account_id)
JOIN customer_segments USING (account_id)
WHERE fiscal_qtd = 'current'
AND fiscal_day_index <= current_fiscal_day_index
AND segment = 'Enterprise'
GROUP BY lifecycle_stage, lifecycle_range_status;Switch between the dashboard and PPTX preview from one clean output page. Friday keeps the story, metrics, and source trace attached.
Revenue is down 9.0% versus previous fiscal QTD. Enterprise is the only statistically unusual drop, and the top 8 Enterprise accounts explain 61% of the gap.
See how Friday keeps context from question to answer.
Private beta · Free during beta · No credit card · Local-first
Friday is built for analytics teams that cannot hand over their data estate. It runs locally, uses your model key, and does not retain or train on your work.
Memory, artifacts, and settings live in Friday Desktop.
Local-first, BYOK, zero retention. Join the private beta.
Private beta · Free during beta · No credit card · Local-first
Friday learns schemas, table metadata, business rules, and project decisions as you work, turning them into memory that improves every analysis.
Always compare customer segments as they were at quarter start.
Fiscal QTD uses the company fiscal calendar.
Revenue uses net_revenue_usd after refunds and credits.
Renewal-stage splits come from subscriptions.lifecycle_stage, not revenue_events alone.
QTD outlier checks use 2σ against each segment's trailing 12-quarter movement.
Pilot accounts tagged is_pilot=true stay out of booked revenue unless finance approves.
Top-account Pareto cuts rank by absolute QTD revenue gap, not percent change.
Friday remembers schemas, rules, and findings as you work. Join the private beta.
Private beta · Free during beta · No credit card · Local-first
Friday is the workspace you see; Fridaybot is the engine underneath. It plans analysis, writes code, builds artifacts, and keeps schema, memory, metric rules, and source trace attached to every run. Model-agnostic and BYOK, so you can bring your own provider key.
Model-agnostic harness, your keys, your data. Join the private beta.
Private beta · Free during beta · No credit card · Local-first
No — you drive Friday in plain English. But it's an IDE, not a black box: every line of SQL it generates is there to read, run, and verify, with the reasoning and source trace attached. You don't have to write the code, but you can always check it.
Friday connects to warehouses like BigQuery and Snowflake, plus files and apps. Point it at a table and it learns the schema and metadata automatically, so analysis runs on governed, trusted data without heavy upfront setup.
Point Friday at a new table and it auto-learns the schema, metadata, and meaning. Metric definitions, business rules, and findings then accumulate as governed memory you control — review, edit, approve — and every new question starts by gathering the relevant slice instead of you re-explaining it each time.
Yes, both. Friday keeps the generated code, reasoning, selected tables, assumptions, and source trace attached to every artifact, so you can inspect the whole path from question to chart. And any analysis can be published as a dashboard or PPTX deck with that context still attached — so you share the answer without losing how it was produced.
Yes. Friday runs locally on your desktop, uses your own model key (BYOK), and never retains or trains on your prompts, data, or outputs.
BYOK means bring your own key — you connect your own model-provider key, so usage stays under your account and Friday never owns your prompts, data, or outputs. Fridaybot is model-agnostic, so you can route work through providers like Anthropic, OpenAI, Gemini, or local models as support is added — the same analytics context and workflow either way.
Fridaybot is the analytics harness under Friday. It plans the next step in an analysis, calls the model you connect, generates SQL and artifacts, and keeps schema, memory, KPI rules, and source trace attached to every run.
Beta users get 3 months free after public launch. After that, founder pricing locks in at $17/month — the rate you keep as an early user.
Full workspace access: governed context, chat, generated SQL, pinned artifacts, dashboards, and PPTX export — all with source trace still attached.
A coding agent lives in a repo and produces patches. Friday lives around business questions and produces verified analysis. That's what "IDE for analytics" means: just as an IDE keeps code, context, execution, and output together, Friday keeps your question, governed context, generated SQL, checks, artifacts, and final story in one traceable workspace.
Friday is built for data professionals — analysts, analytics engineers, and data scientists. You don't need to be a software engineer, but it's designed for people who think in questions and data and want to inspect the work, not just receive a chatbot answer.
Friday is built for the messy work before and around BI: exploring a question, writing analysis code, checking assumptions, pinning artifacts, and packaging the result. Published dashboards complement the BI tools your team already uses.
You get the full workspace free. Your feedback shapes what we build next.
Private beta · Free during beta · No credit card · Local-first