Scaling Uber with Thuan Pham (Uber's first CTO)
The Pragmatic Engineer · video · 2026-04-10 Reader 🖥️
Building and scaling a technology organization under extreme growth pressure requires survival-oriented pragmatism in systems design, relationship-driven talent acquisition, and a leadership philosophy rooted in humility and purpose — not grand strategic planning but iterative responses to existential constraints.
practical · long-form interview / oral history — A deep conversational interview with Uber's first CTO covering engineering leadership, scaling systems under extreme growth, and career philosophy. Primarily practical wisdom drawn from lived experience.
Structure
| Part | Role |
|---|---|
| Thuan's origin story: Vietnam, refugee journey, self-taught programming, MIT | Establishes character and foundational values: resilience, resourcefulness, intrinsic motivation |
| Pre-Uber career: HP Labs, SGI, Netgravity, dotcom bust, VMware | Illustrates pattern of deliberately cycling between comfort and discomfort, and key business lessons (timing, market readiness, growth vs. profitability) |
| Joining Uber: Bill Gurley connection, 30-hour interview with Travis | Demonstrates how reputation and relationships create opportunities; shows Travis's engineering-mindedness |
| Scaling through chaos: dispatch rewrite, vertical scaling limits, database failures, monolith decomposition | Core technical narrative — survival-mode engineering under hypergrowth |
| China launch in 5 months | Exemplifies the culture of impossible deadlines producing extraordinary results; introduces the 'hardest first' principle |
| Microservices and program/platform split | Explains how organizational and architectural decisions emerged from necessity, not ideology |
| Project Helix (app rewrite), naming culture, level system, internal transfers | Illustrates organizational maturation challenges as company scales |
| Three tours of duty and departure | Framework for how leaders decide when to stay and when to leave |
| Post-Uber: Coupang, Nubank, Fair, AI and the future | Validates the transferability of principles; introduces current thinking on AI and engineering |
Key terms
- brick wall / runway
The point at which a system will catastrophically fail under growing load, and the time remaining before hitting it - program and platform
Uber's organizational structure: programs are cross-functional teams building end-user features; platforms build tools and infrastructure for program teams - tour of duty Career Development
A self-defined phase of engagement with a company organized around a specific purpose (fix, scale, or stabilize) - redline yourself
Push yourself beyond perceived limits to discover true capability, as Travis encouraged at Uber - see around corners
The CTO/leadership responsibility to anticipate technical and organizational challenges 18-24 months ahead while the team focuses on near-term execution - swarm coding Multi-Agent Orchestration
Orchestrating multiple AI coding agents working in parallel on related tasks, requiring multi-threaded thinking from the human engineer
Propositions
- explicitUnder extreme growth, you must optimize system rewrites for survival runway rather than architectural perfection — build only what buys you the next 12 months.conf: high
- explicitCareer opportunities are primarily generated by reputation compounding through doing excellent work and treating people well, not by deliberate networking.conf: high
- explicitUber's thousands of microservices emerged from survival necessity under hypergrowth, not from an architectural philosophy or preference.conf: high
- explicitWhen facing an impossible deadline with multiple components, doing the hardest part first produces better outcomes than incremental risk management.conf: high
- explicitThe traits that make a great engineer — curiosity, fearlessness, willingness to innovate — remain constant regardless of how much tooling changes, including AI.conf: high
- explicitMaking internal transfers easy is strategically superior to restrictive processes because engineers who want to leave will interview externally anyway.conf: high
- explicitGreat technology alone is insufficient — commercial viability and market timing are equally necessary for success.conf: high
- explicitcorroboratedA CTO's most important jobs are (1) building a high-performance team with talent density and cultural alignment, and (2) seeing 18-24 months ahead while the team handles near-term execution.conf: highsupports ←Netflix CPTO on AI and the future of product and tech roles | Elizabeth Stone · p12 — Both name talent density as the foundational, non-negotiable input a leader must secure before anything else (agency, risk-taking, forward planning) can work.
- implicitDeliberately oscillating between large comfortable organizations and small risky ones is a valid strategy for continuous professional growth.conf: high
- implicitThe people you bring with you from previous roles are as important to your success as your own individual capabilities.conf: high
Arguments
Survival-oriented engineering — not architectural purism — is the correct approach under hypergrowth. strength: high
Explicit premises:
- Uber's dispatch system would hit a brick wall in 5 months if not rewritten
- A perfect rewrite would take a year; a survival rewrite took 3 months
- The monolith decomposition (Darwin) that should have taken 3-6 months took 2 years because business growth kept adding to it faster than it could be decomposed
Implicit premises:
- A dead company has no architecture to improve
- The rate of business growth, not engineering preference, should determine architectural decisions
Inferences:
- Build only what buys you 12 months of runway, then reassess
- Systems will need to be rewritten multiple times — this is expected, not failure
Microservice proliferation at Uber was an emergent survival strategy, not an ideological choice, and should not be emulated without similar growth pressure. strength: high
Explicit premises:
- The monolith prevented teams from moving independently
- New features had to be built outside the monolith as microservices to avoid blocking
- Even while decomposing the monolith, other teams were adding to it faster
- 10 years later, Uber has fewer microservices than at peak
Implicit premises:
- Most companies do not experience Uber-level growth violence
- Microservices impose coordination costs that are only justified by extreme parallelization needs
Inferences:
- Unless you're growing at the speed of Uber, you probably will not need thousands of microservices
Reputation is the primary mechanism for career advancement at the senior/executive level. strength: high
Explicit premises:
- Bill Gurley remembered Thuan from Netgravity a decade earlier
- VMware engineers followed Thuan to Uber because they enjoyed working with him
- Some people worked with Thuan across 5 companies over 28 years
Implicit premises:
- Senior hires happen through trusted referral networks, not job boards
- The willingness of top talent to follow a leader is a stronger signal than any interview
Inferences:
- Investing in relationships through genuine excellence is a compounding asset
Narratives
- Thuan's family fled Vietnam as boat people with less than 50% survival odds, arrived in the US with no English and no money, sponsored by a church with donated clothing.
→ Establishes resilience and resourcefulness as foundational character traits; contextualizes all subsequent risk-taking as calibrated against genuine existential risk - The dispatch rewrite race: 5 months to rewrite a single-threaded NodeJS dispatch system before New York City traffic killed it, with only two constraints (N boxes per city, M cities per box).
→ Exemplifies the survival-oriented engineering philosophy — minimal requirements, maximum urgency, leading questions rather than directives - China launch in 5 months: Travis wanted 2 months, industry benchmarks said 18 months, they negotiated to 4, slipped to 5, launched Chengdu (hardest city) first.
→ Illustrates the 'hardest first' principle and how impossible deadlines can produce extraordinary results when teams commit fully - Thuan begged on LinkedIn for anyone with Postgres kernel expertise to help diagnose random failures, and could not find help at any price.
→ Motivates the decision to build internal infrastructure (schemaless, M3, etc.) — the terror of depending on open-source software with no commercial support at extreme scale - VMware engineers from Denmark followed Thuan to Uber and built schemaless and core infrastructure — every one he called came, despite asking 'why a taxi company?'
→ Demonstrates that reputation and relationships are the mechanism through which leaders actually execute — you cannot build alone
Questions addressed
- ✅ Why did Uber end up with thousands of microservices?
Survival necessity under hypergrowth: new features had to be built outside the monolith to avoid blocking, while decomposition of the monolith (Darwin project) took 2 years because the business was growing faster than code could be extracted. - ✅ Why did Uber build so many internal tools instead of using existing open-source solutions?
They started with open-source (Redis, Postgres, etc.) but pushed each to breaking point at Uber's scale. The Postgres incident — where no expert could be found at any price — crystallized the need to control their own infrastructure destiny. - ✅ How should a CTO decide when to leave?
Three criteria: (1) Do I love the mission? (2) Am I making a big impact? (3) Am I enjoying the people? When several dimensions are lacking and money is not a factor, it is time to go. - ⭕ How will AI change software engineering?
AI is already dramatically increasing output for top engineers (2x+ at Fair). The current frontier is applying AI to build new features on top of large existing codebases rather than greenfield projects. The traits that make great engineers remain constant. - ✅ What advice for young engineers entering the industry now?
Three career phases: (1) First 5-10 years — maximize learning and skill development. (2) Senior/staff level — seek maximum impact, take risks on smaller companies. (3) Principal/VP level — give back, coach, develop others.
Gaps
- unstated_assumptionThe interview assumes that Uber's survival-mode engineering culture was net positive, but does not examine the human cost (burnout, 'people went to the beach and stared at water for a month' after China launch). The sustainability of this approach is not questioned.
Read next: Examine research on sustainable pace in engineering organizations; compare with less intense scaling stories (e.g., Stripe, Shopify) that grew large without the same crisis-driven culture. - missing_perspectiveThe narrative is told entirely from the CTO's vantage point. The experience of mid-level and junior engineers living through constant rewrites, impossible deadlines, and organizational churn is not represented except as positive growth.
Read next: Seek accounts from IC engineers who worked at Uber during 2013-2017; Gergely Orosz's own newsletter has some of these but they are not surfaced here. - unexamined_tensionThuan advocates for both 'redlining yourself' (pushing past limits) and humility/perspective-of-death (don't take yourself too seriously). How these two values interact — especially when the 'redlining' culture contributes to the very organizational crises (2017 turbulence) he later had to stabilize — is not explored.
Read next: Literature on organizational culture tensions: Edgar Schein's work on espoused values vs. enacted values; research on high-performance cultures that avoid toxicity. - survivorship_biasThe framing 'we did the impossible and it worked' does not account for the many startups that adopted similar crisis-driven engineering and failed. Uber's extreme approach worked partly because Uber succeeded as a business — the engineering culture was necessary but not sufficient.
Read next: Case studies of companies with similar engineering intensity that did not succeed (e.g., WeWork's engineering org, Theranos's move-fast culture). - underdeveloped_topicThe 2017 turbulence (Travis's departure, leadership crisis) is mentioned but deliberately not explored. This is arguably the most consequential organizational event in Uber's history and Thuan was at the center of it.
Read next: Mike Isaac's 'Super Pumped'; public reporting on Uber's 2017 leadership crisis.