The corporate chapter is closed. The 2nd innings starts now.
22 years — Microsoft, Google, Hiya, Agoda, Binance, N26, Bybit — done. What follows is not retirement. It's a different game with a different scoreboard.
4 yrs
The window before Ryshi leaves home
~27 hrs
Building per week, school hours only
87 kg
Target weight by mid-2027 (from 102)
$100K+
Eventual annual venture income goal
The one-paragraph version. The scarce resource for the next four years is not money — it is Ryshi's remaining time at home and my health span. Run a balanced week: 25–30 hours of building during school hours, everything else protected. Attack the weight with medical help plus environment design, not willpower. Validate RetireOdds with customers before writing another line of code. And consciously fire myself from the CPO role at home. Income is a real goal — on a multi-year clock that must never outrank the window.
Operating Principles
Four rules everything answers to
1 The four-year window is the deadline
Ryshi is 14. Bangkok is home until he leaves for college. Every venture decision gets tested against: does this cost me the window?
2 Solo + AI, by design
No employees, no cofounder, no investors. The leverage is 22 years of product judgment multiplied by Claude. It also means nobody else does the customer conversations.
3 Income is a goal, not a deadline
Target ~$100K+/yr from ventures eventually — ~1,000–1,700 subscribers at solo-product pricing. No burn, no runway pressure, no reason to rush a bad bet.
4 Health is the keystone habit
Patience with the kids, energy for Mona, stamina for building — all downstream of the body. Health goes first in the day, not last.
The Default Week
Built around a late start & the markets
~25 hours of building with health and family structurally protected — not fit in around work. The day starts ~08:30 and ends with a two-hour capital-markets block. Sunday evening: a 30-minute review against the scoreboard. I've run biweekly CEO product reviews for years; this is that, for my life.
Time
Mon–Fri
Weekend
08:30–09:00
Slow start, breakfast
Family breakfast
09:00–10:30
Health block with Mona — strength 3×/wk, zone-2 walk or swim 2×/wk
Long walk or pickleball together
10:30–15:30
Build block (RetireOdds) — deep work, phone away
Sat: 1:1 time, kid's choice of agenda
15:30 →
Kids' time — pickup, robotics, math, hanging out
Mnemara3d sessions, family time, trips
Evening
Family dinner, no screens at the table
Fri 20:00 pickleball · one date night per week
22:00–24:00
Markets block — trading / monitoring. Ends at midnight sharp; no eating during it
Optional — reading & research
Two rules on the markets block: midnight is a hard stop (lights out ~00:30 protects the 8 hours of sleep the weight plan depends on), and trading is portfolio management — it doesn't count toward the 25–30 build hours.
The Four Pillars
Where the effort actually goes
1 · Health — 102 kg → 86–88 kg
Food is the fight. Mona is a co-founder, not a supporter.
BMI 28.6 today. Target ~BMI 24 by mid-2027 — 0.35–0.5 kg per week. Slow is correct; crash pace loses muscle and rebounds.
GLP-1 consult in the first two weeks — Bumrungrad or Samitivej. Baseline labs plus a DEXA scan so we track fat vs. muscle, not just kilograms.
Strength is non-negotiable on GLP-1 — 3×/week resistance training, ~150g protein/day, or the drug takes muscle along with fat.
Fix the environment upstream — set the cook's menu weekly; the default becomes compliant without daily willpower. Restaurants: pre-decide the order, protein first.
One KPI: 7-day average weight, logged weekly. No elaborate dashboards in month one — the habit matters more than the tooling.
2 · RetireOdds
Prototype built. Zero users. The risk isn't building — it's hiding in building.
Rule #1 — No new features until 20 customer conversations are done.
The open question: my network is HNW (Long Angle), my chosen market is mass-affluent pre-retirees. Two different businesses — content/SEO play at $50–100/yr, or premium DIY tier at $500–2,000/yr. Talk to 10 of each; let the conversations decide.
Weeks 1–4: 20 problem conversations — how do they answer "can I retire?" today, what do they pay for, what scares them.
Weeks 5–8: landing page live, prototype in 10 hands, and charge something. Payment is the only validation signal that counts. Exit: ≥3 people pay real money.
Weeks 9–12: double down on whatever produced paying users — or write an honest one-page kill/pivot memo.
Day-90 gate: if nobody pays, that's a cheap fast answer — bet #2 gets the same 90-day treatment. Hypothesis pipeline, not identity.
3 · Ryshi (14) & Arya (11)
The failure mode has a name: executive brain at home.
The KPI for parenting is not throughput. It's: do they still seek me out at 18? Every optimizing intervention that wins the moment and erodes that metric is a bad trade.
Invert the decision-velocity rule at home. At work, nothing waits 24 hours. At home, almost every correction can — and half dissolve if you let them.
In Mnemara3d I am the investor, not the CPO. They own decisions, deadlines, mistakes, and profits. Board seat: questions, not directives. Shipping something worse than I would have — that's the curriculum.
1:1s where they set the agenda. Weekly slot for each kid. No progress reviews, no IGCSE talk unless they raise it.
The "family business" stays an aspiration for now. This summer: a real micro-business loop — make, sell to strangers, handle money, keep profits. The inheritable asset isn't a company; it's the capability and the taste for building.
4 · Mona
The health project is the shared project.
Joint by default: the morning block, the menu, the weekly weigh-in. Build the streak together.
One protected date night per week — the first thing the CPO calendar always ate.
Convert time-freedom into experiences: a couples trip every quarter — Bangkok's 2–3 hour flight radius makes it easy.
One conversation to have soon: what does she want from her chapter? The plan has a slot reserved for her answer.
Guardrails
The four failure modes to watch
The gravitational pull of the old life.
Advisory offers, board seats, "one small engagement." Rule: 90-day no-commitment window. After that, anything recurring must fit inside the 25–30 hr build budget and beat RetireOdds for the slot.
Building tools instead of businesses.
MCP servers, dashboards, trackers — beautiful tooling when the real task is uncomfortable. Tooling counts as build time only if a customer will feel it.
Solo isolation.
From daily CEO syncs to a laptop. One peer conversation per week — Long Angle, Skip, builder friends. Loneliness shows up as depression or unnecessary hiring; pre-empt both.
Identity vacuum.
"CPO of Bybit" answered "what do you do?" for years. The new answer, ready and comfortable: "I build small products solo, and I'm raising two teenagers." Both halves are the job now.