Professional services

Help when you want it, not as a condition of buying.

The product is self-serve. These engagements exist for teams who want a hand — a fractional C-level seat, platform work that stands on its own, or help with a migration, a pilot or a platform review around the gateway.

Fractional leadership

The seat, without the full-time hire.

Chief AI, Technology, Architect and Information Officer roles, taken as a fractional seat — with the implementation services below to back it, so the strategy and the build do not drift apart. These are seats we have held, not a service line we invented, and the person who writes the strategy stays accountable for what it costs to run.

Chief AI Officer

An AI strategy that survives contact with your data estate: what to build, what to buy, what to refuse — and the governance to say why. The implementation hands come with the seat, not as a second engagement.

Chief Technology Officer

Technology direction for the stretch where you are too big to wing it and too small for the full-time hire. Roadmap, build-versus-buy, vendor leverage, and the engineering standards to hold all of it.

Chief Architect

The boundaries between systems, drawn by someone who stays accountable to them. Reference architecture sized to the team you actually have, and the authority to say no to the pieces you do not need.

Chief Information Officer

The information estate run as a part-time seat with full-time accountability: budget, risk, procurement, security posture and the data organization itself.

Around your platform

The practice that existed before the product did.

We spent our careers inside enterprise data platforms — Databricks, Airflow, Lakehouse builds, the API and policy layer in front of all of it. That work is still a business, it does not require you to run anything of ours, and it is not a channel for selling you the gateway. If an engagement would go better without us in the path, that is what we will tell you.

Assessment

Read the estate as it actually is — what runs, what it costs, what nobody owns any more — and write down a sequence worth doing. The deliverable is a decision, not a diagram.

Architecture and design

Lakehouse, warehouse, orchestration and the boundaries between them. Designed so the next team can hold it, which usually means fewer pieces than the reference architecture suggests.

Implementation

Databricks, Airflow and the integration work between them, built with your engineers rather than beside them. We write code on these engagements; we do not just review it.

Enablement

Your team runs it when we leave, or the engagement failed. Pairing and documentation are part of the scope, not a phase that gets cut when the schedule slips.

Optimization and governance

Cost, performance and data quality, with the controls to keep them where you put them. This is where the same instinct behind the product shows up: the cheapest work is the work that stops happening.

Where it overlaps

Warehouse spend comes up in nearly every assessment, and the free scan answers a useful part of it in an afternoon. Run it before you talk to us — it makes the first conversation concrete and it costs you nothing.

Around the gateway

Scoped to Airbrx: get behind it, prove it, tune it.

Migration

Moving traffic behind the gateway across more than one tool, engine or team — and deciding what order to do it in.

Integration itself is a hostname change. This is for the coordination around it, not the technical work.

Pilot

A 30-day POC run with you against a target agreed up front, so what decides it is a number from your own engine.

Included with the Enterprise POC — see pricing.

Platform review

Reading your query history with you: what repeats, what would cache well, what has been asking your data, and what to write rules for first.

Start with the free scan — it answers a good part of this on its own.

How engagements are scoped

Fixed scope, named outcome, and an end date.

  1. 1

    Say what would count as working

    Before scope, before price. For gateway work the free scan usually is that number, and it costs nothing to run. If the outcome cannot be stated as something measurable on your side, that is worth knowing before anyone signs.

  2. 2

    Scope to that, and nothing adjacent

    Fixed scope against that outcome. Anything worth doing that turns up mid-engagement gets its own scope and its own decision — not quietly folded into this one.

  3. 3

    Finish, hand over, and stop

    Engagements end. Nothing here is a retainer you have to unwind, and none of it is a condition of buying the product — which is self-serve and free to start.

The track record is the people — Apigee through Google’s acquisition, a WaveQ exit, the CTO seat at M Science, Silicon Valley Bank, Wine.com, KPIT, Birlasoft and NWEA — and it is on the about page under the names it belongs to.

You do not need any of this to use Airbrx. The product is self-serve, Starter carries no platform fee, and integration is a connection string. Services exist for teams who want a hand — not as a gate in front of the product.

Bring a problem, or bring a number.

Either works. Run the scan and tell us what you want to move, or bring the platform problem as it stands — it is a better first conversation than any deck we could send you.