Timeline
10-12 hours
PRO · module 01 free previewStaff Engineering trackP18
Lead like a
Staff+
engineer: RFC → ADR → Review → Postmortem
Four simulations that produce the artifacts a staff+ promo committee actually reads. Author a Netflix-format RFC for a streaming migration. Build an ADR catalog with a 5-year TCO. Facilitate an architecture review with documented dissent. Write a Google-SRE blameless postmortem for a 6h data-loss incident. No code shipped — every deliverable is the kind of writing a calibration committee reads.
The staff+ loop at Stripe, Meta, Google, Uber asks you to walk through a design you led, defend the tradeoffs, and explain what you learned from the last incident. This project gives you all three — already written, already pressure-tested.
By the end you will have authored
- RFC-042-streaming-migration.md — Netflix-format design doc with 3 scored alternatives, Mermaid data-flow + sequence diagrams, and a phased rollout timeline
- ADR catalog — 5+ decisions with weighted tradeoff matrices and a 5-year TCO model (Kafka vs Kinesis vs status-quo batch)
- Architecture-review packet — timed agenda, RACI, disagree-and-commit framework, escalation playbook, consensus document with recorded dissent
- INC-2024-0312 postmortem — Google SRE format, 5 Whys + fishbone + contributing-factor analysis, SLO-driven action items, executive briefing
- Objection playbook — top 10 anticipated questions from VP / CTO / IC with rebuttals keyed to each stakeholder's frame
- Promo-packet bundle — all four artifacts as a polished portfolio, ready to attach to a staff+ promo or interview loop
PREREQBuilt for senior engineers preparing for staff+ scope. You’ve shipped pipelines and led at least one design discussion. Pairs well with the Technical Leadership curriculum — that course teaches the skills, this project produces the artifacts.
promo-packet/ · 4 artifacts · 65+ deliverables
no code · writing onlyProblem → RFC
Tradeoff → ADR
Review → consensus
Incident → postmortem
22h fraud lag$47k/mo · 23 fails/qtr
rfc-042.md
tradeoff-matrix.yaml
diagrams/data-flow.mmd
risk-register.json
scoping · framing
Kafka vs Kinesisvs status-quo batch
adr-001-kafka.md
tco-5yr-model.yaml
objection-playbook.md
decision-log.yaml
5 ADRs · TCO
5 stakeholdersSr. Staff · PM · Plat · DE
review-agenda.md
raci-matrix.yaml
consensus-doc.md
escalation-playbook.yaml
RACI + dissent
INC-2024-03126h data loss · schema break
postmortem.md
five-whys.md
slo-definitions.yaml
executive-summary.md
SRE format
# promo-packet/ (TOC)
├─ 01-rfc-streaming-migration.md
├─ 02-adr-catalog/ (5 ADRs + TCO)
├─ 03-review-packet/ (RACI + dissent)
└─ 04-postmortem/ (SRE format)
● no compile · no infra · plain text
$ unzip staff-engineer-playbook.zip
$ ls templates/ examples/ scenarios/
$ code rfc-042-streaming-migration.md
→ markdown · yaml · mermaid · json
65+
leadership artifacts
4
simulations
10-12h
end to end
Why this matters in 2026
Staff+ promo is judged on writing, not commits.
The senior→staff jump is gated on artifacts a committee can read: RFCs, ADRs, reviews you facilitated, postmortems you owned. Most engineers don't have these — they have a list of features. This project closes that gap with realistic, polished deliverables you can put in front of a calibration committee.
RFC > tickets
A staff+ engineer's promo packet leads with a design doc, not a Jira link. RFC-042 here follows Netflix / Uber / Google formats — the ones promo committees recognize.
ADRs > tribal knowledge
5-year TCO with sensitivity analysis is what stops the same Kafka-vs-Kinesis debate from being re-litigated every quarter. Reversibility tiers + recorded dissent are how you institutionalize the decision.
Facilitation > consensus theatre
Architecture reviews fail because they lack a frame. RACI + disagree-and-commit + an escalation SLA is what makes them produce decisions instead of meetings.
Blameless postmortems > 'don't do that again'
Google SRE format + prevention hierarchy (eliminate → automate → guardrail → document) is the difference between learning from the incident and re-running it next quarter.
Curriculum · 4 modules · 10-12 hours
Module 01 is free. The rest unlocks with PRO.
Try the first 2.5-3 hours — author a real RFC for a streaming migration with scope, tradeoff matrix, Mermaid diagrams, and a phased timeline. If the rhythm clicks, upgrade to unlock the ADR catalog, facilitated review, and postmortem modules.
P18 · 10-12 hours · 4 modules
Free preview EXPERT required
Module 01 is free — no card required. Author the RFC and feel the writing standard before paying.
M01
✓Author an RFC: streaming-migration design doc
Write a Netflix-format RFC for a daily-batch → Kafka/Flink streaming migration at a fintech (22h fraud-detection latency, 23 pipeline failures/quarter, $47k/mo Snowflake spend). Quantified problem statement, 3+ alternatives with weighted tradeoff matrix, Mermaid data-flow + sequence diagrams, risk register, phased timeline with rollback criteria.
M02
⊘Defend the choice: ADR catalog + 5-year TCO + objection playbook
Build the ADR catalog around the streaming-migration choice. Score Kafka vs Kinesis vs status-quo on a weighted matrix. Model 5-year TCO (infra + personnel + opportunity cost) with sensitivity on traffic and retention. Write per-frame messaging — VP hears cost/risk, CTO hears tech depth, IC hears career impact. Top-10 objection playbook with pre-written rebuttals.
M03
⊘Facilitate the review: agenda + RACI + dissent
Run the architecture review for the same streaming migration with 5 stakeholders (Sr. Staff pushing Kinesis on cost, PM pushing Q2 deadline, Platform worried about ops burden, Data Eng wanting schema-evolution guarantees, you facilitating). Timed agenda, RACI for decision ownership, disagree-and-commit protocol, escalation playbook with 24-48h SLA, consensus doc with recorded dissent, action-item tracker.
M04
⊘Own the incident: blameless postmortem + executive briefing
Reconstruct INC-2024-0312 — payments team deploys breaking schema change, Flink consumer silently drops 1.84M messages over 6h, dbt freshness test catches it at T+6h13m. Build the timeline from logs, run 5 Whys + fishbone, classify contributing factors (caused / prolonged / amplified), write the Google-SRE format postmortem, set SLO-driven action items, and brief leadership in 5 slides.
3 modules locked · Unlock all PRO content for $29/mo
Upgrade to PRO →Backed by curriculum
Staff Data Engineer: Leadership & Architecture
Technical Leadership is the curriculum this project is the capstone for — that course teaches the skills, this project produces the promo-packet artifacts. PRO subscribers get full access to every module.
The build, in 3 phases
Three sprints. Three checkpoints. One promo-ready packet.
Each phase ships named artifacts you can attach to a real promo or interview loop. No theory decks.
01~3h
RFC + scope + tradeoff matrix
Netflix-format design doc for the streaming migration with quantified problem, 3 scored alternatives, Mermaid diagrams, risk register, and phased timeline.
- ✓rfc-042-streaming-migration.md
- ✓tradeoff-matrix.yaml
- ✓risk-register.json + diagrams/
02~5h
ADR catalog + TCO + facilitated review
Defend the choice with a weighted ADR catalog and 5-year TCO model. Run the architecture review with timed agenda, RACI, and recorded dissent.
- ✓adr-001-kafka.md + tco-5yr-model.yaml
- ✓objection-playbook.md (top 10)
- ✓review-agenda.md + raci-matrix.yaml + consensus-doc.md
03~3h
Postmortem + executive briefing
Reconstruct INC-2024-0312 from logs, run 5 Whys + fishbone, write the Google-SRE blameless postmortem with SLO-driven action items, brief leadership in 5 slides.
- ✓INC-2024-0312-postmortem.md
- ✓slo-definitions.yaml + action-items.yaml
- ✓executive-summary.md (5-slide briefing)
Project setup · 5 minutes
One zip. Templates, worked examples, and a real incident scenario.
No Docker, no cloud credentials, no compile step. Unzip the kit, open it in your editor, and start writing. Every artifact is plain text — Markdown, YAML, Mermaid, JSON.
What lives in the kit
Three reusable templates plus a fully worked end-to-end case study you can read before you write your own. The incident scenario is the raw input for the Part 4 postmortem exercise.
- templates/ — 3 reusable templates: RFC (Netflix format), ADR catalog, Google-SRE postmortem
- examples/part1-design-doc/ — full worked RFC with scope, tradeoffs, Mermaid diagrams, risk register
- examples/part2-adr-catalog/ — ADRs, 5-year TCO model, objection playbook, decision log
- examples/part3-architecture-review/ — agenda, RACI, escalation playbook, consensus doc with recorded dissent
- examples/part4-postmortem/ — Google-SRE postmortem, 5 Whys, SLO action items, exec briefing
- scenarios/incident_scenario.json — raw 6h analytics data-loss incident (schema-change → silent drop) for Part 4
Download · Starter Kit
Staff+ Leadership Playbook Starter Kit
Three reusable templates, 63 worked-example artifacts across all 4 parts, and the raw incident scenario. All YAML / JSON / Mermaid / Markdown lints cleanly out of the zip.
~/projects/staff-engineer-playbook — zsh
1. Unzip the kit and open it in your editor
$ unzip staff-engineer-playbook-starter.zip
$ cd staff-engineer-playbook-starter && code .
2. Browse the reusable templates (RFC, ADR catalog, postmortem)
$ ls templates/
$ # design_doc_template.md adr_catalog_template.yaml postmortem_template.md
3. Read the worked end-to-end case study (streaming-migration RFC → review → postmortem)
$ ls examples/
$ # part1-design-doc/ part2-adr-catalog/ part3-architecture-review/ part4-postmortem/
4. Use the incident scenario as the raw input for the Part 4 postmortem
$ cat scenarios/incident_scenario.json | jq '.incident.title'
$ # "6h analytics data loss due to upstream Kafka schema change"
5. (Optional) Lint the YAML / JSON / Mermaid before adapting to your team
$ find . -name '*.yaml' | xargs -I {} python3 -c "import yaml; yaml.safe_load(open('{}'))"
$ find . -name '*.json' | xargs -I {} python3 -c "import json; json.load(open('{}'))"
3
reusable templates
63
worked artifacts
1
incident scenario
4
parts wired
Senior-grade vs staff-grade
Same artifact. The difference between senior and staff is in the rigor.
Most engineers can write something called “an RFC” or “a postmortem.” The promo-committee bar is higher. Each row below pairs the senior-grade version most engineers ship today with the staff-grade upgrade you produce in this project.
Senior-grade artifactWhat most engineers ship
×
Risks
Bullet list at the bottom of the RFC
×
Cost claim
“Kafka is cheaper” — no model
×
Alignment
Verbal “sounds good” in Slack
×
Pushback
One-stakeholder pitch, same deck for everyone
×
Postmortem
“5 Whys” ending in a name
×
Followup
“Make sure it doesn't happen again”
Staff-grade artifactModule 01–04
✓
Risks
risk-register.json with likelihood × impact, mitigation owners, rollback triggers✓
Cost claim
tco-5yr-model.yaml — 5-year TCO with sensitivity on traffic + retention assumptions✓
Alignment
adr-001.md — weighted scorecard, recorded dissent, reversibility tier (Type 1 vs Type 2)✓
Pushback
objection-playbook.md — per-frame messaging: VP hears cost/risk, CTO hears tech depth, IC hears career impact✓
Postmortem
5 Whys + fishbone + “where we got lucky” — system-level, not person-level
✓
Followup
Prevention hierarchy:
eliminate → automate → guardrail → document — tied to SLOs and error budgetsPRO benefit · written-artifact review
Real review from staff+ engineers who’ve sat on promo committees.
Submit your RFC, ADR, or postmortem. Get line-by-line feedback within 48 hours from engineers who've actually been the calibration reviewer. The kind of feedback that's hard to get inside one company.
WR
4 artifact reviews / month
Submit an RFC, ADR catalog, or postmortem. Reviewer is matched to your domain — platform / streaming / batch — and has shipped at staff+ at companies that gate on this writing. Async, comments inline, average turnaround 31 hours.
31h
avg turnaround
9.2/10
helpfulness
94%
return next month
OH
2 office hours / month
Live 30-min sessions for promo prep — walk through your packet, mock a calibration discussion, get the 'is this staff-quality?' read before submission. Group sessions for review-meeting practice also available.
30 min
per session
2 / mo
included
+ group
unlimited
What PRO unlocks
One subscription. 15+ projects, all curriculum, written-artifact review.
PRO is built for engineers preparing for staff+ scope — production builds, leadership artifacts, and feedback loops. Not more tutorials.
What you getFREEPROEXPERT
Projects
Production-grade builds
2
15+
8
Curriculum modules
All 7 tracks
Phase 1 only
All
All + bonus
Code / artifact review
Senior + staff engineer review
0
4 / month
Unlimited
Career path access
5 paths × full plans
1 path
All 5
All 5 + 1:1
Certificate
Verifiable on LinkedIn
—
Yes
Yes + portfolio review
Community
Discord + office hours
Read-only
Full + 2/mo
Full + 4/mo
$29/mo
billed monthly · cancel anytime
or annual
$249/yr save 28%
Who this is for
Pick this if you’re going up for staff+, not learning to code.
SD
Senior DE → Staff
You have the technical chops; you need the artifacts a promo committee can actually read. This produces them in the formats committees recognize.
TL
Tech leads
You run reviews already; you want a repeatable framework for facilitation, RACI, escalation, and recording dissent without breaking trust.
EM
EMs with IC roots
You're calibrating staff candidates on your team. This is the rubric for what staff-quality writing looks like — useful for self-assessment and for mentoring.
AR
Architects / principal candidates
You make decisions across teams but document them in Slack. This forces those decisions into ADRs and postmortems with the rigor a principal review expects.
Related curriculum
Going deeper? Three tracks back this playbook.
This project is the artifact-production capstone. These three curriculums let you go deeper on the underlying skills — the leadership rituals, the system-design content, and the metrics-and-stakeholder vocabulary every staff+ artifact assumes.
FAQ
Quick answers.
Because staff+ promotion isn't judged on code; it's judged on writing, tradeoffs, and how you handle disagreement. Every artifact in this project is the kind of writing a promo committee actually reads — RFC, ADR, RACI, blameless postmortem. Hands-on means you write them, not that you compile something.
System Design teaches the *content* of distributed systems (ingestion, storage, serving, scaling). This project teaches the *artifacts* and *process* around any decision — how to write an RFC, defend a tradeoff, run a review, own a postmortem. Most senior+ engineers need both, but they fail promo on this one.
Technical Leadership teaches the skills (8 modules — RFCs, design reviews, cross-team ownership, vision). This project is the capstone where you produce the artifacts those skills generate, on a single end-to-end case study (the streaming-migration scenario). Pair them: curriculum first if you want depth on the skills, project first if you want to produce the deliverables fast.
No. Every deliverable is plain text — Markdown, YAML, Mermaid, JSON. The starter kit is a 57 KB zip; you unzip it, open in your editor, and start writing. No compile step, no infra.
Yes. RFC follows Netflix / Uber / Google patterns; ADRs follow the Michael Nygard format; postmortem follows Google SRE; RACI is the industry-standard decision-ownership matrix. These are the formats promo committees and architecture review boards already use.
All 4 modules of this project, all 15+ PRO projects, 4 written-artifact reviews per month, 2 office-hours sessions, full curriculum across all 7 tracks, all 5 career paths, certificate of completion, and full community access. Cancel anytime.
Ready to build a promo-ready packet?
Start with module 01 — free, no card. About 2.5-3 hours. By the end you'll have a Netflix-format RFC for a streaming migration with quantified scope, three scored alternatives, Mermaid diagrams, and a phased timeline — the kind of artifact that leads a staff+ promo packet.
P18 · Staff+ Leadership Playbook · PRO · module 01 freeUpgrade to PRO →