About | Oliver Cheng (Oli)

while (world) { observe(); infer(); hallucinate?(); self-correct(); } while (world) { observe(); infer(); hallucinate?(); self-correct(); } while (world) { observe(); infer(); hallucinate?(); self-correct(); }

001101 010011 sigil::bagua seed::entropy trace::mechanical awe::human 001101 010011 sigil::bagua seed::entropy trace::mechanical awe::human 001101 010011 sigil::bagua seed::entropy trace::mechanical awe::human

temperature=0.92 top_p=0.95 sampler=stochastic runtime=deterministic user=astonished temperature=0.92 top_p=0.95 sampler=stochastic runtime=deterministic user=astonished temperature=0.92 top_p=0.95 sampler=stochastic runtime=deterministic user=astonished

☰☱☲☳☴☵☶☷ oracle != truth ritual == interface meaning <- interpretation ☰☱☲☳☴☵☶☷ oracle != truth ritual == interface meaning <- interpretation ☰☱☲☳☴☵☶☷ oracle != truth ritual == interface meaning <- interpretation

seed = hash(question + state); noise = sample(temperature); pattern = deterministic(seed, noise); if (human_can_track === false) mark("mystic"); return explainability_gap(pattern); seed = hash(question + state); noise = sample(temperature); pattern = deterministic(seed, noise); if (human_can_track === false) mark("mystic"); return explainability_gap(pattern);

## Not a person, still persuasive - interface implies intention - language implies confidence - user infers agency => design for interpretability ## Not a person, still persuasive - interface implies intention - language implies confidence - user infers agency => design for interpretability

cast.bagua = pickTrigrams(seed); cast.moonBlocks = deriveHexagram(seed); cast.fortuneSticks = burnModel(intensity); cast.scapula = generateCracks(seed); return readable_fiction(cast); cast.bagua = pickTrigrams(seed); cast.moonBlocks = deriveHexagram(seed); cast.fortuneSticks = burnModel(intensity); cast.scapula = generateCracks(seed); return readable_fiction(cast);

{ "observe": true, "decide": constrained, "act": reversible, "measure": behavior, "learn": weekly } { "observe": true, "decide": constrained, "act": reversible, "measure": behavior, "learn": weekly }

def perceivable_randomness(system): return complexity(system) > attention_budget if perceivable_randomness(llm): user.labels_output = "fate" def perceivable_randomness(system): return complexity(system) > attention_budget if perceivable_randomness(llm): user.labels_output = "fate"

[trace] t=02:13 system murmurs [trace] tokens fall like ash [trace] certainty simulated [trace] mechanism remains [trace] human names it chance [trace] t=02:13 system murmurs [trace] tokens fall like ash [trace] certainty simulated [trace] mechanism remains [trace] human names it chance

Profile

About Oliver Cheng (Oli)

I build AI products across PM, UX, and engineering, and I work closely with teams to get them shipped.

I work at the product-engineering boundary. I take work from ambiguous problem to shipped product, then run the iteration loop based on real usage and feedback with the team.

I am the founder of Bonzen, an AI coaching and micro-meditation product. Previous roles include product, design, and engineering work at Unyte, SmartHalo, Esper, and IBM.

I can lead a workstream end to end, but I also interweave with existing team processes: align cross-functional stakeholders, guide scope decisions, unblock implementation details, and keep delivery moving.

Based in San Francisco
Works async + live collaboration
Builds for web + mobile contexts
Comfortable in 0->1 and team-scale ambiguity

Current

Thought of the Month

February 2026 AI note.

February 28, 2026

Agents Need Self-Modeling Humans

AI agents are most useful when the operator understands their own goals first. Treat each agent as an autonomous extension of a specific objective, define the task and constraints clearly, and keep enough technical literacy to verify what the agent is actually doing.

Clarity before delegation.
One agent per concrete goal when possible.
Audit outputs with technical judgment, not trust theater.

Read full thought

Experience

Selected Background

Timeline across product, design, and engineering roles.

2024 - Present

Independent

Remote (US/Canada)
Building & Consulting

Building Bonzen and multiple shipped web apps while advising startups on AI product execution. Day-to-day work includes product framing, UX, implementation, launch, and post-launch iteration.
- AI products
- iOS
- Web apps
- LLMs
- 0->1 execution
2020 - 2024

Unyte

Remote (US/Canada)
Product Manager

Led new product development from concept through release. Owned roadmap decisions, prototypes, user research, and cross-functional execution with engineering, design, sales, and leadership.
- SaaS
- User research
- Roadmapping
- Cross-functional execution
2018 - 2020

Aterica

Waterloo
Product Designer

Designed digital health product experiences connected to medical hardware. Focused on UX systems, usability testing, and translating clinical constraints into usable interfaces.
- Digital health
- UX/UI
- Design systems
- Medical device context
2016 - 2017

SmartHalo

Montreal
Product Designer + Engineer

Designed and built companion experiences for a connected cycling product. Ran user experiments, improved mobile flows, and worked across hardware, firmware, and app teams.
- IoT
- Mobile apps
- Hardware-software UX
- Experimentation
Early career

Esper

San Francisco Bay Area
Software Engineer (Co-op)

Built onboarding flows at an early-stage startup working on ML-assisted workflow automation. Established a foundation in fast product iteration and startup execution.
- ML workflows
- SaaS
- Onboarding
- Startup pace

How I Work

Operating Principles

Standards I use to collaborate with teams, reduce ambiguity, and keep delivery moving.

Align collaboratively

Create shared clarity across product, design, and engineering before execution.
Ship iteratively

Ship small complete increments with clear ownership and visible progress.
Guide with evidence

Use product behavior data to guide decisions, coaching, and next-step prioritization.

Start a conversation PM Resume Engineer Resume

About Oliver Cheng (Oli)

Thought of the Month

Agents Need Self-Modeling Humans

Selected Background

Operating Principles

Align collaboratively

Ship iteratively

Guide with evidence