back
Author models

Frameworks

Ideas from 100+ talks in one visual form. Each one-pager reads in 30 seconds and ships with a single link.

Topics
01

CTO Lifecycle

Coder → Delivery → All-in-one → ?
Leadership
When to use
Use when aligning which CTO profile the company needs at its current growth stage.
Company lifecycle →↑ CTO scopeStartupGrowthMaturityNext betCoderbuilds the productDeliveryships at scaleAll-in-onetech + people + business?platform · CPTO · new S-curve

The CTO role evolves with the company lifecycle. At founding the CTO is the main coder; in growth — owns delivery; at maturity — becomes an all-in-one leader spanning technology, people and business.

02

Channel → Product → Platform

Three-step evolution of a large fintech
PlatformArchitecture
When to use
Use when product and platform strategies start to conflict because they are in different phases.
01 · 2006 – 2013
Channel
website as a shop window
02 · 2014 – 2020
Product
ecosystem around the client
03 · 2021 – now
Platform
shared capabilities for many teams
2006
first site
2010
online bank
2014
mobile · super-app
2017
lifestyle services
2020
internal platforms
2022
platform contracts

Channel — the website as a bank shop window (2006). Product — an ecosystem of services around the customer (2014). Platform — shared capabilities reused by dozens of teams and partners (2022).

03

SE 2.0 → SE 3.0

Code-first → Intent-first engineering
AI-nativeArchitecture
When to use
Use when a team moves from local copilots to agentic workflows and specifications.
A
SE 2.0 · code-first
B
SE 3.0 · intent-first
Code
→ Main artifact →
Intent + spec
Engineer writes lines
→ Author →
Engineer writes prompts & specs
Code review + tests
→ Quality gate →
Evals + contracts + sampling
Copilot autocomplete
→ Speed lever →
Agentic workflows
Typing & reading code
→ Bottleneck →
Context, intent, evals
Algorithms & frameworks
→ Skill that wins →
Problem framing & taste

SE 2.0 — a world where the main artifact is code. Tests, reviews and CI all serve it. Copilots speed up such an engineer 1.5–2×.

04

Senior+ Fork

Team track vs Tech track
Leadership
When to use
Use when an engineer chooses the next step after Senior and the company designs a dual ladder.
Senior Engineer
owns own delivery
Team track
results through people
L1Teamlead
L2Engineering Manager
L3Director
L4VP / CTO
Tech track
results through architecture
L1Staff
L2Principal
L3Distinguished
L4Fellow

Past Senior the career forks into two full-fledged tracks: a people track (Teamlead → Engineering Manager → Director → CTO) and a tech track (Staff → Principal → Distinguished → Fellow).

05

Fitness Functions Matrix

Atomic ↔ Holistic × Triggered ↔ Continuous
ArchitectureSRE
When to use
Use when architectural qualities need automated checks instead of team memory.
Atomic
single service / module
Holistic
the whole system
Triggered
on event / in CI

Unit-level checks in CI

  • ArchUnit — no cyclic deps in a module
  • Lint rules on a service
  • Per-service contract tests

System checks on demand

  • Load test of an end-to-end flow
  • Chaos game-day for a domain
  • Security pentest on release
Continuous
always-on in production

Always-on per-service signals

  • Service SLO + error budget
  • p99 latency alerts
  • Crash-free sessions for a mobile app

Always-on system-wide signals

  • End-to-end business SLOs
  • Domain dashboards & alerts
  • Cost / resource budgets

Fitness functions from evolutionary architecture are automated checks that the system preserves its important characteristics over time. The 2×2 matrix lays them out along two axes.

06

Dream Teamlead AI-native

Individual → Team → Org culture
LeadershipAI-native
When to use
Use when a teamlead's AI habits need to become a shared team contract.
L3Org culture
shapes the system
AI quality metrics
Security & privacy policy
Sharing rituals
Hiring for the new profile
L2Team effect
shared contract
Shared prompts & evals
Agentic delivery flows
Shared telemetry
AI onboarding
L1Individual effect
personal productivity
AI for code & review
Notes & planning agents
Search & research
Drafting & comms

An AI-native teamlead works on three levels. Individual — personal productivity and AI tooling for routine: code, reviews, notes, planning.

07

SRE Maturity

From on-call heroes to error budgets and chaos
SREArchitecture
When to use
Use when you need to assess reliability maturity and choose the next 2-3 practices.
L1Reactive
What we measure:uptime
How we respond:heroics
How we learn:firefighting
L2Managed
What we measure:SLA
How we respond:runbooks
How we learn:postmortems
L3Proactive
What we measure:SLO + error budget
How we respond:automation
How we learn:blameless reviews
L4Resilient
What we measure:user-journey SLO
How we respond:self-healing
How we learn:game-days
L5Antifragile
What we measure:business SLI
How we respond:auto-rollback
How we learn:chaos in prod

SRE maturity is not 'do we have an on-call rotation', but how deeply the discipline is wired into the product. A 5-step ladder: Reactive → Managed → Proactive → Resilient → Antifragile.

08

Engineering Productivity

DORA + DevEx + SPACE — three lenses, one picture
ProductivityLeadership
When to use
Use when engineering productivity needs a metric system rather than a single counter.
DORAlens · 01
delivery flow
  • Deploy frequency
  • Lead time for changes
  • Change failure rate
  • MTTR
↳ How fast & safe do we ship?
DevExlens · 02
engineer experience
  • Flow state
  • Feedback loops
  • Cognitive load
↳ How does it feel to work here?
SPACElens · 03
outcome & context
  • Satisfaction & wellbeing
  • Performance
  • Activity
  • Communication
  • Efficiency & flow
↳ Are we actually productive?
⚠Goodhart's law: optimize one counter alone — and the other two collapse. Read all three together.

Measuring engineering needs three complementary metric sets. DORA — delivery flow (deploy frequency, lead time, change failure rate, MTTR). DevEx — engineer experience (flow, feedback loops, cognitive load).