Use this when a CTO or engineering leader needs to plan technology investments over the next 6-18 months, connecting business objectives to infrastructure, platform, and tooling decisions. Also use when presenting a technology strategy to the board or executive team.

Related skills: Use /tech-debt-assessment for current-state input. Use /now-next-later-roadmap for the formatting pattern. Use /migration-plan for items requiring migration. Use /decision-brief for investment approval. Use /annual-strategy-review for strategic alignment. Use /build-vs-buy for individual technology decisions.

Process

Step 1: Gather inputs

Collect the following from the engineering leader and cross-functional stakeholders:

Business strategy/objectives -- what the company is trying to achieve in the next 12-18 months
Growth projections -- expected user/revenue/data growth rates
Product roadmap -- committed and planned product initiatives
Known scaling challenges -- where systems are already straining
Regulatory requirements -- upcoming compliance deadlines (SOC 2, GDPR, accessibility)
Current tech debt hotspots -- reference /tech-debt-assessment output if available
Team capabilities and planned hiring -- current headcount, skill gaps, hiring plan
Vendor contracts approaching renewal -- contracts expiring in the roadmap window

Step 2: Inventory current technology landscape

Map the current state across these dimensions:

Systems in production (services, databases, queues, caches)
Languages, frameworks, and major libraries
Infrastructure providers and regions
Key external dependencies (APIs, SaaS tools)
End-of-life or end-of-support status for critical components
Known performance bottlenecks and reliability gaps

If a /tech-debt-assessment exists, use it as the primary input here rather than duplicating the analysis.

Step 3: Map technology initiatives to business objectives

For each business objective from Step 1, identify the required technology capabilities. For each capability, document:

Field	Detail
Business objective	Which objective this serves
Capability needed	What the technology must do
Current state	What exists today
Gap	What needs to change
Effort estimate	Team-months of engineering work
Infrastructure cost impact	Net change to monthly infrastructure spend
Dependencies	Other initiatives or external factors this depends on

Initiatives that don't connect to a business objective need a clear rationale (foundational work, risk reduction, compliance) or they don't belong on the roadmap.

Step 4: Prioritize and sequence

Apply the Now/Next/Later framework:

Now (0-3 months) -- committed work, teams assigned, requirements clear
Next (3-9 months) -- planned work, scoped but not yet staffed, dependencies identified
Later (9-18 months) -- directional bets, needs validation before committing

Sequencing considerations:

Dependencies between initiatives (what must finish before what can start)
Team capacity constraints (don't plan for 150% utilization)
External deadlines (compliance dates, contract renewals, vendor EOL)
Quick wins that build credibility for larger investments

Step 5: Generate the roadmap document

# Technology Roadmap -- {{time_period}}

## Executive summary
{{2-3 sentences connecting total technology investment to business value}}

## Investment overview
- Total engineering investment: {{X}} team-months
- Infrastructure cost change: {{+/- $X/month}}
- Vendor/licensing changes: {{summary}}
- Key decision points: {{dates when commitments must be made}}

## Now (0-3 months) -- Committed

### {{Initiative name}}
- **Business objective:** {{which business goal this serves}}
- **Current state:** {{what exists today}}
- **Target state:** {{what we're building toward}}
- **Investment:** {{team-months, infrastructure cost}}
- **Timeline:** {{start -- end}}
- **Risks:** {{top 1-2 risks}}
- **Owner:** {{team or person}}

## Next (3-9 months) -- Planned
{{Same card format, with note that scope may adjust}}

## Later (9-18 months) -- Directional
{{Same card format, with explicit uncertainty markers}}

## Dependency map
{{Which initiatives depend on others, critical path}}

## Sunset plan
{{Technologies being retired, migration timelines, cost savings}}

## Appendix: Cost of doing nothing
{{For each major initiative, what happens if we don't invest}}

Step 6: Produce executive presentation summary

Create a one-page summary for board or executive consumption:

# Technology Strategy Summary -- {{time_period}}

**Total investment:** {{X}} team-months + {{$X}} infrastructure/vendor
**Expected business impact:** {{2-3 bullet points connecting to revenue/growth/risk}}

**Top 3 initiatives:**
1. {{Initiative}} -- {{one-line business impact}}
2. {{Initiative}} -- {{one-line business impact}}
3. {{Initiative}} -- {{one-line business impact}}

**Top 3 risks:**
1. {{Risk}} -- {{mitigation}}
2. {{Risk}} -- {{mitigation}}
3. {{Risk}} -- {{mitigation}}

**Key decisions needed:**
- {{Decision}} by {{date}}
- {{Decision}} by {{date}}

Step 7: Review

Before finalizing, pressure-test the roadmap:

Does the sequencing account for real team capacity (not theoretical)?
Are there technology bets that need validation (POC, spike) before committing resources?
Which initiatives have external dependencies (vendor timelines, regulatory deadlines)?
Is the sunset plan realistic, or are we assuming migrations happen faster than they do?
What's the cost of doing nothing for each major initiative?
Will the executive audience understand why each investment matters?

Output location

Save the roadmap document to deliverables/technology-roadmap-{{date}}.md or the client's preferred location. The executive summary can be a separate file or a section within the roadmap.

Example Output

Input

Company: Meridian Health Analytics — Series C SaaS company providing clinical decision support tools to mid-size hospital networks
Business objectives: Expand from 40 to 120 hospital clients by end of 2026, launch an AI-powered risk scoring module in Q3 2025, achieve SOC 2 Type II certification by June 2025
Growth projections: 3× patient data volume (1.2TB → 3.6TB/month ingested), API call volume growing 250% as new EHR integrations go live
Known scaling challenges: Single-region PostgreSQL cluster at 78% CPU during peak hours; monolithic Rails app deploy times averaging 42 minutes; no staging environment for enterprise clients
Team: 18 engineers today, hiring 6 more by Q2 2025; gap in ML Ops and infrastructure/platform engineering

Output (abbreviated)

Technology Roadmap — Q1 2025 through Q2 2026

Executive Summary

Meridian's technology investments over the next 15 months are organized around three business imperatives: surviving 3× data growth without service degradation, accelerating the AI risk scoring launch that anchors the Series C narrative, and closing enterprise deals that require SOC 2 Type II. The roadmap commits 74 team-months of engineering effort and adds approximately $38K/month in net infrastructure spend, offset by $14K/month in retiring legacy vendor contracts.

Investment Overview

Dimension	Detail
Total engineering investment	74 team-months
Net infrastructure cost change	+$24K/month (+$38K new, −$14K sunset)
Vendor/licensing changes	Retire Heroku standard dynos; add Snowflake enterprise tier; renegotiate PagerDuty at renewal in March 2025
Key decision points	Database migration architecture by Feb 1 · MLOps platform (build vs. buy) by March 15 · SOC 2 auditor selection by Jan 15

Initiative Map to Business Objectives

Initiative	Business Objective	Gap	Effort	Infra Cost Impact
PostgreSQL → Aurora multi-region	3× growth + 99.9% uptime SLA for enterprise	Single-region, 78% peak CPU	10 team-months	+$12K/month
Rails monolith decomposition (Phase 1)	Faster product iteration for AI module	42-min deploys block weekly releases	14 team-months	+$4K/month
SOC 2 Type II compliance program	Enterprise sales requirement	No formal controls program exists	8 team-months	+$3K/month (tooling)
ML risk scoring infrastructure	AI module launch Q3 2025	No model serving or feature store	16 team-months	+$11K/month
EHR integration platform (HL7 FHIR hub)	Scale to 120 clients (most require Epic/Cerner)	Point-to-point integrations, 6-week lead time per client	12 team-months	+$8K/month
Heroku → AWS ECS migration	Prerequisite for all scaling work	Heroku limits multi-region and custom networking	8 team-months	−$14K/month (savings)
Developer platform (staging envs, CI/CD)	Engineering velocity for growth hiring	No per-client staging; onboarding new engineers takes 3 weeks	6 team-months	Neutral

Now (0–3 months) — Committed

Heroku → AWS ECS Migration

Business objective: Foundational — prerequisite for database, ML, and integration work
Current state: All services running on Heroku standard dynos; no VPC, no private networking
Target state: Core services containerized and running on ECS Fargate in us-east-1 with private subnets, secrets in AWS Secrets Manager
Investment: 8 team-months · saves $14K/month vs. Heroku at current scale
Timeline: Jan 6 – Mar 28, 2025
Risks: Deploy pipeline regression during cutover; Heroku add-on dependencies (Heroku Redis, Heroku Postgres) must migrate in parallel
Owner: Platform team (2 engineers, Infrastructure lead)

SOC 2 Type II — Controls Foundation

Business objective: Enterprise sales — 6 of 12 pipeline deals require SOC 2 before signing
Current state: No formal access controls documentation, no audit logging, no vendor risk management process
Target state: Controls framework live, audit logging enabled across all services, auditor engaged for Type II observation period starting April 2025
Investment: 3 team-months engineering + $18K auditor fee · $3K/month tooling (Vanta)
Timeline: Jan 6 – Mar 31, 2025 (observation period April–June)
Risks: Audit findings in April may require additional engineering remediation sprint
Owner: Head of Engineering + designated Security Lead (new hire, starting Feb)

Next (3–9 months) — Planned

PostgreSQL → Amazon Aurora (Multi-Region Read Replicas)

Business objective: Support 3× data volume growth and enterprise uptime SLAs
Current state: Single RDS PostgreSQL instance, us-east-1 only, 78% CPU at peak ingestion
Target state: Aurora cluster with writer in us-east-1, read replica in us-west-2, autoscaling read capacity, query optimization pass on top 20 slow queries
Investment: 10 team-months · +$12K/month
Timeline: Apr – Aug 2025
Dependencies: ECS migration must be complete (networking prerequisites)
Risks: Schema migration for 1.2TB live database requires zero-downtime strategy; budget approval for Aurora pricing increase
Owner: Data infrastructure squad

ML Risk Scoring Infrastructure

Business objective: Q3 2025 AI module launch — primary Series C growth driver
Current state: Data science team running experiments in Jupyter notebooks; no feature store, no model serving layer, no retraining pipeline
Target state: SageMaker-based feature store + model registry, REST inference endpoint (<200ms p99), automated weekly retraining on new patient data
Investment: 16 team-months · +$11K/month
Timeline: Apr – Sep 2025
Dependencies: Aurora migration (feature store needs stable schema); MLOps hire (targeting April start)
Decision needed by March 15: Build feature store on SageMaker vs. buy Tecton or Feast managed offering
Risks: MLOps hiring is on critical path; 6-week slip in hiring = 6-week slip in launch
Owner: ML platform team (new) + Data Science lead

Developer Platform — Staging Environments & CI/CD

Business objective: Engineering velocity to absorb 6 new hires and sustain weekly releases
Current state: Shared staging environment; new engineer onboarding takes ~3 weeks; no per-PR preview environments
Target state: Ephemeral per-PR environments on ECS, GitHub Actions pipeline under 12 minutes, onboarding runbook targeting 3-day time-to-first-commit
Investment: 6 team-months · cost-neutral (reuses ECS infrastructure)
Timeline: Apr – Jun 2025
Owner: Platform team

Later (9–18 months) — Directional

HL7 FHIR Integration Hub

Business objective: Reduce client onboarding from 6 weeks to under 2 weeks; required for 80% of target hospital clients
Direction: Dedicated integration service with pre-built Epic and Cerner connectors, async job queue (SQS), and per-client configuration UI
Investment estimate: 12 team-months · +$8K/month
Timeline: Oct 2025 – Feb 2026
Uncertainty: Scope depends on Epic App Orchard certification timeline (external dependency)
Validation needed: Spike in Q2 2025 to confirm FHIR R4 compatibility with target EHR versions

Rails Monolith Decomposition — Phase 1

Business objective: Release velocity and team autonomy as engineering grows past 24 engineers
Direction: Extract notifications and reporting as independent services; introduce internal API contracts; do not attempt full microservices rewrite

Related Engineering skills

Architecture Discovery Boris Model Event Storm SNAP Document System Diagram Security Review Pinning Tests TDD Tests

Back to Skills Catalog