VECTORIO“For commerce ops teams that decide in seconds but report in hours”
City-scale inventory analytics on one box.
City-level inventory and dispatch analytics on one machine — the same code you run at scale.
- Stock shows available, the picker finds an empty shelf, the order is cancelled.
- Pricing runs overnight, but competitors reprice during the day.
- Every dark store is its own island when the link drops.
- City-level inventory and dispatch analytics on one machine — the same code you run at scale.
- Replay a day to explain why a price or assignment was made.
- Store or hub runs on a local box and syncs when connected.
From today to the first result.
a streaming platform plus overnight batch jobs
Pricing runs overnight while competitors reprice during the day, and a stock count that shows available meets an empty shelf.
Streaming and batch in one engine: 103 million rows through a 21-stage workflow in 24.4 s end to end, and any day can be replayed to explain a price.
Replay yesterday's events through a workflow next to the current pricing job and compare the prices.
one week pilot100 million events, cleaned and trusted. Step by step.
A day of streaming events is checked at the door, pseudonymised, de-duplicated over three days, enriched and rolled up per user — bad rows land in a dead-letter table and stay out of the numbers.
- Streaming Events · Session Context · Tracks Dimension · Schema ConformanceCSV Subscriber · CSV Subscriber · CSV Subscriber · Schema Conform
- Timestamp · PII BoundaryCompute Fields · Key Generator
- Event ID Generator · Event DeduplicationKey Generator · Deduplication
- Rollup Column Prep · Session Enrichment · Track Enrichment · Daily User RollupCompute Fields · Lookup · Lookup · Rollup Aggregation
- Derived Columns · Quality Gate · Daily Rollup OutputCompute Fields · Sync.QualityGate · CSV Publisher
- Output Fork · Trusted Events Output · Quality Statistics · Quality Metrics Output · DLQ Error Handler · DLQ OutputFork · CSV Publisher · Window Functions · CSV Publisher · DLQ Handler · CSV Publisher
event-source → CSV.Subscriber
session-context-source → CSV.Subscriber
tracks-dim-source → CSV.Subscriber
schema-conform → ETL.SchemaConform
timestamp-normalization → ETL.Compute
pii-boundary → ETL.KeyGen
event-key-gen → ETL.KeyGen
dedup → ETL.Dedup
rollup-prep → ETL.Compute
session-enrichment → ETL.Lookup
track-enrichment → ETL.Lookup
derived-columns → ETL.Compute
quality-gate → Sync.QualityGate
output-fork → ETL.Fork
trusted-output → CSV.Publisher
daily-rollup → ETL.Rollup
rollup-output → CSV.Publisher
quality-stats → ETL.Window
quality-output → CSV.Publisher
dlq-handler → ETL.DLQHandler
dlq-output → CSV.Publisher- Streaming EventsReads the 100M-row listening-event file in 131,072-row batches with auditing on.
- Schema ConformanceForces required fields and safe types so bad rows are caught at the door.
- TimestampTurns the timestamp text into epoch milliseconds for time logic.
- PII BoundaryReplaces the user id with a keyed HMAC-SHA256 pseudonym; the raw id stays inside this step.
- Event DeduplicationDrops duplicate events with a Bloom filter over a sliding 3-day window (capacity 120M, 0.1% false positives).
- Session EnrichmentAdds user tier, subscription and device type by session id.
- Daily User RollupAggregates plays per user per day; closes days using an event-time watermark with 72h allowed lateness.
- Quality GateChecks every row against rules and routes failing rows to their own output, where they stay available.
- Daily Rollup OutputWrites the daily per-user totals.
A real workflow from our end-to-end test suite; the designer view is illustrated.
What you get. From the first pilot.
City-level inventory and dispatch analytics on one machine — the same code you run at scale.
Replay a day to explain why a price or assignment was made.
Store or hub runs on a local box and syncs when connected.
Why your tools fall short.
Cloud streaming stacks are costly and complex at modest scale.
Store-level edge cases (offline, intermittent links) are not handled by central warehouses.
Batch tools are too slow for per-minute decisions.
- Margin pressure in quick commerce
- Stockout and cancellation costs
- Competitive repricing speed
- Peak events (sales days) that break pipelines
Your pilot. Three steps.
Your data shape
Tell us what data you have, where it lives and what the result must be.
A scoped pilot
On your data and your hardware, with the success measure written down first.
A clear decision
You see the result against that measure — and keep every workflow and record.
Related roles in Cost & reproducibility.
Data/analytics engineering teams fighting cloud warehouse bills
Stop paying a cluster for laptop-sized queries.
Finance & trading risk analytics
Risk numbers you can reproduce on demand.
System integrators and consultancies building data platforms for clients
One engine to deliver, audit and hand over.
ISVs and SaaS vendors needing an embedded analytics / ETL engine
Built into your product.
AI/ML teams
Reproducible data prep at memory speed.
Insurance analytics
Reproducible reserves, run in minutes.
Mid-size companies without a data team
Replace the spreadsheet chain, keep the data.
Bring us your hardest data problem. Let's solve it together.
A short pilot on your data and your hardware. We agree on what success looks like before we start.