BitscorpSKU · 4 weeks or retainer
Data Engineering
One pipeline from your events to a warehouse you own, plus a dashboard of metrics you can defend.
Four weeks to stand up one named pipeline — Firebase, ads, Mixpanel, S3, or Postgres — into BigQuery or Postgres, with freshness checks and a dashboard of the metrics you freeze in week 1. €12,000 (50% up front) or €3,000/mo with a three-month minimum.
The wound
What this is for
DAU lives in three tools and none of them agree. Monday is a CSV. Attribution is a screenshot. You cannot answer which campaign paid back, which cohort stayed, or whether the last release moved retention. A lakehouse programme is a year. A spine — one pipeline, one warehouse, ten metrics — is a month.
- A mobile, SaaS, or marketplace team with events in Firebase, ads, Mixpanel, or S3 and no warehouse they trust
- A founder who can name 5–10 metrics that would change a decision this quarter — not “we need a data team”
- Someone who will give source access in week 1 and will not ask us to invent the product strategy from a dashboard
Data Engineering (4 weeks)
€12,000
50% to start, 50% at week-4 demo. Written SOW required.
Analytics retainer
€3,000/mo
Monthly in advance, three-month minimum. Pipeline + dashboard upkeep for the frozen metric list, not a blank cheque for new domains.
Cap: two concurrent retainers studio-wide. Written SOW. No handshake starts.
Four-week Data Engineering
What the weeks look like
Week 1
Sources, contracts, metric freeze
Inventory sources, access, and grain. Freeze the metric list in writing (DAU, retention, LTV, campaign, or the five you actually use). If we cannot reach the sources or you cannot name the metrics in five days, we stop. You keep the deposit unused beyond that week’s work.
Week 2
Warehouse and pipeline
Stand up BigQuery or Postgres as the warehouse of record. Orchestrate with Airflow or Luigi. Partition, idempotent loads, and a freshness check — not a pile of notebooks.
Week 3
Dashboard and quality
One dashboard you own (Metabase, Looker Studio, or Grafana). Row counts vs source. Late-data rule. The numbers have to survive a sceptical founder, not a demo.
Week 4
Soak and hand over
Runbook, who owns the DAG, what pages when freshness breaks, and a recorded walkthrough. Optional retain starts here if you want us to keep the spine alive.
In
Included
- One named pipeline (events, ads, tickets, or app store) into BigQuery or Postgres
- Airflow or Luigi DAG with retries and a freshness check
- One dashboard of the week-1 metric list
- Slack, one business-day response on that pipeline
- Weekly demo slot (Tue or Thu)
1 business day on the Data Engineering channel
Out
Not included
- A company-wide lakehouse or “modern data stack” programme
- Training a custom ML model or a recommendation engine
- Rewriting the product, LocaleOps, or AI Consulting
- 24/7 on-call
- Owning your GCP / AWS / Azure account — you invite us; you keep billing
Questions
Before you book
No. Data Engineering is a pipeline plus a dashboard of named metrics. If you want a model, a recommender, or a research intern, this SKU is the wrong buy — we will say so.
Whichever your volume and your team can own. Firebase + Ads usually land in BigQuery. Smaller app or game ledgers often stay in Postgres. We pick in week 1, in writing.
We say so in writing. You do not pay the second 50%. You pay for the week of inventory you already received.
Named SKUs. Pick one. Do not invent a new one on the call.
Data Engineering (4 weeks) €12,000 · Analytics retainer €3,000/mo. Written SOW. 100% up front unless the offer says otherwise.