Roadmap
Framework

Why we run two complementary roadmaps, how they work together, and the principles guiding everything the Data Science Team plans and builds.

01 · Context

Why We Need Two Roadmaps

Our team operates in a rapidly evolving AI/ML landscape where business priorities, operational challenges, and emerging technologies shift quickly. Traditional static roadmaps assume stable requirements — but our inherently exploratory work doesn't. A model experiment can pivot into a production system overnight. A stakeholder need can reframe an entire quarter.

At the same time, we can't operate purely reactively. Without investing in durable capabilities — strong data infrastructure, reusable ML patterns, consistent observability practices — we'd spend every cycle rebuilding foundations instead of compounding on them.

The tension we resolve

Short-term delivery pressure vs. long-term capability investment. We don't choose — we track both, separately, so neither gets lost in the other.

02 · Principles

Three Guiding Principles

Every roadmap decision — what to prioritize, what to defer, what to drop — is measured against these three.

Stay Agile

Adapt initiatives to evolving business needs. AI/ML work is inherently exploratory — priorities shift, experiments pivot, and breakthroughs require rapid re-alignment. Our roadmap is designed to flex without losing direction.

Build Capabilities

Invest in long-term strengths that compound. While initiatives come and go, capabilities persist. Every project should leave the team more capable than before — stronger infrastructure, deeper knowledge, reusable patterns.

Deliver Impact

Convert experiments into scalable value. Research and prototypes matter only when they translate to better decisions, reduced toil, or real business outcomes. Every roadmap item ties back to a measurable impact.

03 · The Two Views

Two Complementary Roadmaps

To balance short-term delivery and long-term growth, we split planning into two distinct views — each with its own rhythm, evidence, and ownership.

View 1

Adaptive Initiatives

What we're building

Tracks active and planned projects based on current business priorities. Initiatives are the execution layer — quarterly-driven, business-aligned, and pivotable as needs evolve. While specific deliverables may shift, every initiative contributes to our long-term capability strategy.

  • Quarterly execution timeline with initiative lifecycle stages
  • Status tracking: Discovery → Exploration → Validation → Operationalization → Scaling
  • Progress visibility per quarter with business impact scoring
Open Adaptive Initiatives

View 2

Capability Development

What we're becoming

Tracks long-term strategic growth across six core competency areas. Unlike initiatives, capabilities focus on building durable competencies, reusable systems, and scalable practices that persist beyond individual projects and survive team changes.

  • Six core capability pillars: Data Engineering, Analytics, ML, GenAI, MLOps, Governance
  • Maturity model from Foundational (L1) through to Leading (L5)
  • Role-based contribution mapping across the team
Open Capability Development
04 · Key Questions

What the Roadmap Answers

Together, the two views answer four core questions stakeholders and the team regularly ask.

Strategic Direction
Where is the team heading long-term?
Initiative Planning
What are we working on now and next?
Team Progress
Where is the team currently at?
Execution Strategy
How are we going to do it?

Two roadmaps. One destination.

Sustainable, compounding impact — delivered by a team that builds as it executes.