🏅 शाळेच्या पद्धतीने engineering leadership शिका

# the 60-second wow — one department, twelve lessons:
git clone https://github.com/BaluRaut/learn-engineering-leadership-school.git && cd learn-engineering-leadership-school
python3 lead/demo.py               # 12 lessons: calendars, feedback, forecasts, debt, metrics, on-call
python3 lead/test_lead.py          # 12 checks across the lessons

तुम्हाला काय दिसायला हवे (छाटलेले — यश असे दिसते):

═══ calendar ═══
── Katrina's 40-hour week (09:00–17:00, half-hour slots) · 'focus' = free time in blocks of 2 h or more
   as a teacher (engineer)      meetings  5.0 h · free 35.0 h · focus 30.5 h in  6 blocks · fragments  4.5 h
   as head of maths, scattered  meetings 12.5 h · free 27.5 h · focus 14.5 h in  6 blocks · fragments 13.0 h
   as head of maths, clustered  meetings 12.5 h · free 27.5 h · focus 25.0 h in 10 blocks · fragments  2.5 h
   the lead's meetings by kind: standup 2.5 h · 1:1 2.5 h · planning 1 h · hiring 2 h · sync 1 h · review 2 h · roadmap 1.5 h
   same 12.5 meeting hours, moved to the afternoons → the fragments become mornings you can think in
   the job moved: fewer hours making things yourself, more hours making the department work — protect some maker time on purpose

═══ oneonones ═══
── feedback as a judgement → missing ['situation', 'impact'] · judgement words ['attitude', 'never']
── feedback as SBI        → missing nothing · judgement words none
   SBI names a moment, an observable behaviour and its effect — something Dipika can check, and change
── weekly 1:1 tracker, 12 weeks:
   Dipika     held 11/12 · longest run missed 1 · weeks since last 0
   Aishwarya  held 12/12 · longest run missed 0 · weeks since last 0
   Meera      held 10/12 · longest run missed 1 · weeks since last 0
   Leela      held  5/12 · longest run missed 4 · weeks since last 4  ← overdue: talk this week
   the quietest person, newest to the team, is the one whose 1:1s get moved — the tracker makes that visible

═══ delegation ═══
── who can do each job without help (level 2+ of 3):
   gradebook  3 · Katrina, Dipika, Aishwarya
   release    1 · Katrina
   database   2 · Katrina, Aishwarya
   frontend   3 · Katrina, Dipika, Meera
   runbooks   1 · Katrina
   bus factor 1 · if Katrina is out: ['release', 'runbooks'] have nobody
── after a quarter of pairing (Dipika runs releases, Aishwarya owns runbooks, Meera learns the database) → bus factor 2
── RACI check: weekly release: 2 accountable (Katrina, Dipika) · weekly release: nobody responsible · gradebook migration: 0 accountable (nobody)
   delegation level for 'which test framework the gradebook uses': 7 delegate
   delegation level for 'the exam-week release freeze': 4 agree
   delegation level for 'the hiring bar': 3 consult
   delegate the outcome and the authority, not just the task — and say which level you mean

═══ hiring ═══
── one candidate, rubric of 4 competencies (anchored 1–4), bar 3.0 · scores written BEFORE the debrief:
   independent: Katrina 3.75 · Dipika 2.25 · Aishwarya 2.50 → panel 2.83 → no hire
   after Katrina speaks first (others move halfway to her): Katrina 3.75 · Dipika 3.00 · Aishwarya 3.12 → panel 3.29 → hire
   same evidence, a different decision — written independent scores keep each interviewer's evidence in the room
── calibration on 5 past panels: Katrina +0.37 · Dipika -0.33 · Aishwarya -0.03
   candidate X (Katrina + Aishwarya): raw 3.25 → calibrated 3.08
   candidate Y (Dipika + Aishwarya): raw 3.00 → calibrated 3.18
   different panels, different habits: a lenient panel can outrank a stronger candidate — 5 panels is a small sample, so treat offsets as a prompt, not a correction

═══ forecast ═══
── the last 10 weeks' throughput (items finished): [2, 6, 1, 4, 7, 0, 5, 3, 6, 3] · average 3.7/week · backlog 40 items
   the single-date guess: 40 ÷ 3.7 ≈ 11 weeks
   Monte Carlo, 10,000 seeded trials replaying random past weeks → 50%: 11 weeks · 85%: 13 weeks · 95%: 15 weeks
   chance of finishing within the guess (11 weeks): 58%
── the backlog grows 20% (40 → 48 items) → 50%: 13 weeks · 85%: 16 weeks
   give a range with a probability, and re-forecast every week as the history and the backlog change

═══ priorities ═══
── RICE = reach × impact × confidence ÷ effort (reach: people a quarter · effort: person-months)
   exam results page       1500 × 1 × 80% ÷ 2 =   600
   parent portal login     2000 × 2 × 50% ÷ 4 =   500
   gradebook export fix     400 × 1 × 80% ÷ 1 =   320
   timetable rewrite        300 × 3 × 50% ÷ 6 =    75
   raise the portal's confidence from 50% to 80% → 800, and it jumps to first — the ranking is only as good as that guess
── cost of delay (value lost per week) and duration (weeks): gradebook export fix 5/1 · exam results page 6/2 · parent portal login 8/4 · timetable rewrite 3/6
   biggest value first: parent portal login → exam results page → gradebook export fix → timetable rewrite → total delay cost 142
   CD3 (cost of delay ÷ duration): gradebook export fix → exam results page → parent portal login → timetable rewrite → total delay cost 118
   write the numbers down so people argue with the assumptions, not with each other

═══ debt ═══
── technical debt as a portfolio: principal = hours to fix · interest = hours lost every week (a flat-interest model)
   horizon 26 weeks:
     flaky test suite         principal  24 h · interest 6 h/wk · pays back in   4.0 wk · net  +132 h → pay
     manual release steps     principal  40 h · interest 5 h/wk · pays back in   8.0 wk · net   +90 h → pay
     tangled gradebook core   principal 200 h · interest 4 h/wk · pays back in  50.0 wk · net   -96 h → carry
     old report module        principal 120 h · interest 1 h/wk · pays back in 120.0 wk · net   -94 h → carry
   horizon 104 weeks:
     flaky test suite         principal  24 h · interest 6 h/wk · pays back in   4.0 wk · net  +600 h → pay
     manual release steps     principal  40 h · interest 5 h/wk · pays back in   8.0 wk · net  +480 h → pay
     tangled gradebook core   principal 200 h · interest 4 h/wk · pays back in  50.0 wk · net  +216 h → pay
     old report module        principal 120 h · interest 1 h/wk · pays back in 120.0 wk · net   -16 h → carry
   total interest 16 h/week out of a team's 160 h/week (10%) — paid every week, whether or not anyone decides to
   pay the high-interest, low-principal debt first; carry debt in code nobody touches; re-price when plans change

═══ metrics ═══
── 4 weeks of deploys (the four keys, plus reliability):
   deployment frequency 4.5/week · lead time for changes (median) 18 h · change failure rate 16.7% · failed-deploy recovery (median) 45 min · availability 99.52%
── target 'change failure rate under 10%' → two failures relabelled 'planned hotfix' → CFR 5.6% · availability still 99.52%
── target 'deploy more' → 18 config-only no-op deploys added → frequency 9.0/week · median lead time 1.8 h · users see nothing new
   Goodhart's law, in Marilyn Strathern's words: when a measure becomes a target, it ceases to be a good measure
── flow: cycle times [5, 8, 2, 12, 3, 10] days · flow efficiency 40% (the rest is waiting) — use metrics to ask questions of the system, never to rank people

═══ topologies ═══
── one team of 12: 66 communication paths · two teams of 6: 30 paths inside teams, plus one agreed interface between them
   component teams (frontend/backend/data/platform)   team-waits 14 · features needing >1 team 5/6
   stream-aligned teams + a platform team             team-waits  9 · features needing >1 team 3/6
   stream-aligned + a SELF-SERVICE platform           team-waits  5 · features needing >1 team 0/6
   Conway's law: the system's shape copies the org's communication shape — so design the teams for the system you want
   cognitive load (simple 1 · complicated 2 · complex 3): grades team         4 across 2 domains
   cognitive load (simple 1 · complicated 2 · complex 3): 'everything' team  10 across 5 domains
   team types: stream-aligned · platform · enabling · complicated-subsystem

═══ incidents ═══
── roll back last night's gradebook release     two-way door → decide now, the owner decides, log it, set a revisit date
── try a new standup format for 2 weeks         two-way door → decide now, the owner decides, log it, set a revisit date
── move the gradebook from MySQL to PostgreSQL  one-way door → slow down: write it up, consult, name a decider and a deadline
── sign a 3-year contract for a marking tool    one-way door → slow down: write it up, consult, name a decider and a deadline
   decision log: 3 entries · due for review by week 8: ['new standup format', 'freeze releases in exam week']
── incident, mitigate first (roll back)    detected after 6 min · users hurt for 15 min
── incident, debug forward (find and fix)  detected after 6 min · users hurt for 52 min
   roles: Katrina is incident commander (coordinates, does not debug) · Dipika operates · Aishwarya writes the updates
   afterwards: a blameless review — what made the mistake easy, not who made it

═══ writing ═══
── weekly status meeting, 8 people × 60 min → 368 person-hours a year (46 working weeks)
   written status update: 30 min to write + 8 × 5 min to read → 54 person-hours a year — and it can be searched later
   a 30-minute decision meeting of 6 → 3.0 person-hours; worth it when the decision needs a live argument
── Dipika's first design doc has 3 sections · missing: goals, non-goals, options considered, risks, open questions
   options considered and non-goals are where a reviewer finds the problem cheaply — before the code exists
   a status update: what changed · what is at risk · what I need from you (the ask comes first)

═══ growth ═══
── on-call, 13 weeks, rotation of 3: Katrina 5 wk / 7 after-hours pages · Dipika 4 wk / 3 after-hours pages · Aishwarya 4 wk / 9 after-hours pages
── on-call, 13 weeks, rotation of 5: Katrina 3 wk / 5 after-hours pages · Dipika 3 wk / 3 after-hours pages · Aishwarya 3 wk / 6 after-hours pages · Meera 2 wk / 1 after-hours pages · Leela 2 wk / 4 after-hours pages
   weeks with more than 2 after-hours pages land on: ['Katrina', 'Aishwarya'] — a burnout signal to act on, not a badge
── visible opportunities in 6 months: Dipika 5 · Aishwarya 1 · Meera 0 · Leela 0 — sponsorship follows familiarity unless you track it
   career ladder: expectations per level, in writing · psychological safety: people can say 'I broke it' and 'I disagree'
── the whole picture: protect the week → 1:1s and SBI → delegate the outcome → hire with a rubric → forecast in ranges → price the order → carry debt on purpose → measure the system → shape the teams → decide by the door → write it down → keep people growing and rested

✅ done — the department runs

🗺️ संपूर्ण चित्र — एक आकृती, पूर्ण कोर्स

संपूर्ण कोर्स एका कॅनव्हासवर. 4K आवृत्तीसाठी क्लिक करा.

🗓️ भाग 1 — नवे काम (धडे 1–4)

एक git branch = एक कल्पना; branch 04 मध्ये धडे 01–04 आहेत. lead/ चा प्रत्येक run सारखाच असतो — randomness seeded आहे.

🎲 भाग 2 — नियोजन आणि delivery (धडे 5–8)

काम कधी पूर्ण होईल, आधी काय करायचे, कोणते debt फेडायचे, आणि मोजमाप न बिघडवता delivery कशी मोजायची.

🌱 भाग 3 — teams आणि वाढ (धडे 9–12)

Teams ची रचना, कठीण प्रसंग, lead ची पोहोच म्हणून लेखन, आणि पुढच्या वर्षीही निरोगी असलेली team.

📐 धड्यांच्या आकृत्या — क्रमांक अनुसरा

प्रत्येक धडा एका क्रमांकित बॉक्स-आणि-बाण आकृतीत — स्वतंत्र पानावरही.