Response to your RFI — Polaris to Databricks migration
ConfidentialPlainsight
Executive summary · in motion

From Polaris to Databricks

A once-in-a-decade chance to lower run-cost, raise decision-making agility, and build for the agentic era — where AI agents, not only people, consume and reason over your data.

SAP BW/4HANA Databricks Lakehouse
Scroll
01  The shift

One governed entry point. One core platform.

SAP Business Data Cloud becomes the single, governed door to SAP data. Databricks becomes the core platform. Not a lift-and-shift — a rationalisation.

SAP → SAP BDC (governed gateway) and non-SAP direct → Databricks lakehouse → reports & dashboards, exploratory BI, Gen BI & AI.

02  Not a copy of Polaris
0
Polaris objects in scope

These aren’t technical wiring to be ported — they are years of encoded business meaning. The goal is fewer, cleaner, governed data products, not a tidier monolith.

03  The estate shrinks by design

Harden the few. Rationalise the rest.

We score every object on value, usage, complexity and dependencies — then decide its shape. Harden the mission-critical few as rigid, reconciled products; build durable data products for the core; serve the long tail flexibly; and retire what’s unused.

About a quarter of objects are retired, not rebuilt — and an ad-hoc question becomes a governed pipeline only once it’s frequent and certified. The estate stops growing by default.

04  Two Plainsight accelerators

We compress the heavy lifting.

Accelerator 01

Databricks Rollout Planner

Turns rollout decisions into a governed foundation — workspaces, Unity Catalog, Terraform IaC, CI/CD and a DORA / GDPR / NIS2 / ISO 27001 compliance map — before the first object moves.

Accelerator 02

AI Migration Assistant

A fleet of specialised AI agents that documents the legacy estate, designs the target, then builds and tests on Databricks — while your experts sign off at three gates.

Three human sign-off gates: Document → Design → Build & test. Engineers stay focused on the complex 20%.

05  The outcome

Lower cost. Higher agility.

Fewer hand-built data products, and far less plumbing per question — while faster, context-grounded answers raise decision-making agility. Group Reporting stays a separate, careful track; Polaris remains your system of record until you decide otherwise.

0
Polaris objects in scope
~0 yr
Indicative delivery
0
Base-case total
±0%
Faster, our method

Hands-on engineers, a proven Clear Path method, accelerators for the bulk work, and knowledge transfer so your teams own the result.

Proposal · Confidential

An indicative, part-by-part response to your RFI — our migration approach, target architecture, operating model, governance and pricing for the move from Polaris to Databricks.

Plainsight · Analytics & AI 22 June 2026 Prepared for Atlas Copco Group Support Services NV