Astrodata / Paradero
Confidential · Paradero / Comunal
Your roadmap is right: the data lake is the product, the agent layer arrives faster than any timeline predicts, and data readiness is the only real constraint. Astrodata is the team that turns that architecture into operating reality, with commercial execution as the first and most visible win.
Astrodata builds modern data platforms and the AI capabilities that run on top of them. Our team has shipped these systems in regulated, multi-stakeholder domains where the data layer is the competitive product, not a back-office utility. That is exactly the bet your strategy makes.
We come in as a boutique partner that augments your two-person IT team, takes ownership of the data substrate end-to-end, and makes the hard decisions about identity resolution, retention, and integration architecture so your roadmap stays on schedule.
Server-side tagging via Stape.io into Meta CAPI and Google Ads, a marketing data mart in Fabric, and pace + comp-set forecasting that feeds the Revenue Agent precursor. This is the workstream that shows up first in the P&L: cleaner spend allocation, defensible cost-per-booking, and the attribution substrate that has to exist before Pencil AI or Madgicx are worth evaluating.
Microsoft Fabric Lakehouse architecture, the eleven Azure Data Factory source pipelines, and the identity-resolution layer that joins Opera guest_id with dailypoint contact_id and the WhatsApp phone hash. We design the guest match confidence scoring (HIGH / MEDIUM / LOW), build the manual override workflow for TA and corporate-account guests, and own the LFPDPPP retention rules in Microsoft Purview.
Power BI Semantic Model on Fabric, plus the daily revenue briefing into André's Teams inbox at 7am, which the strategy already names as the first agent deployment. From there: travel-advisor production dashboards (Virtuoso, Signature, Ensemble, Bonotel, TripArc), a homes-buyer pipeline view, group-sales pace, and an owner-statement portal for the rental-pool homes.
Revenue Agent, Pre-Arrival Agent, and Maintenance Agent — built as the strategy specifies, in human-in-loop posture, and only on top of a guest profile that has earned the trust of the team. Pre-arrival drafting goes through dailypoint and Claude API; revenue recommendations go through Duetto and OHIP; maintenance dispatch flows through Quore. The data substrate from Surface 02 is the prerequisite.
We don't build voice agents. We do design the integration architecture that determines whether Riviera, Duve, or any successor writes structured preference tags back to the guest profile, checks real-time OHIP availability, and hands off to a human via Teams Phone with a warm summary already written. Without this layer, the voice agent is an expensive FAQ bot.
guest_uuid prototype with sample dataOn platform alignment: Astrodata's reference stack is Snowflake, dbt Labs, Omni Analytics, and Astronomer. Your strategy commits to Microsoft Fabric for sound consolidation and data-residency reasons. We deliver on the substrate you've chosen — dbt now runs on Fabric Warehouse, ADF and Power Automate are the right integration layers for an OHIP-native hospitality stack, and Power BI is the correct internal analytics tool given M365 leverage. We will tell you when an Astrodata-preferred pattern doesn't apply, rather than steer the architecture toward our defaults.
We'd like 60 minutes with whoever owns the data and commercial calls. Walk through the five surfaces against your real constraints — the dailypoint decision, the Fabric region availability, the IT capacity question — and align on what Phase 01 actually contains before we send formal scope.