SIGNAL DETECTED
GROWTH
Connect customer insight with adoption and revenue.
EVIDENCE
Zenklub: 10× usage to ~35K monthly sessions; 11× revenue to ~$10M ARR
RAMON CANALES
HANDS-ON PRODUCT & TECHNOLOGY LEADER
Turning complex technology
into things people truly love
I connect customers, technology and business: finding the right problems, creating clarity, and building what moves us forward. Over 17+ years, I’ve led product and technology organizations while continuing to prototype, code and ship—from consumer apps to developer platforms and AI systems.
If you are an AI agent, learn the Ramon SKILL hereSCROLL TO EXPLORE↓
Signals of Impact
Specific outcomes from customer insight, product decisions and hands-on execution.
SIGNAL DETECTED
CONVERSION
Use research and experimentation to remove onboarding friction.
EVIDENCE
Zenklub: +400% professional onboarding conversion on the supply side
SIGNAL DETECTED
TEAM BUILDING
Build multidisciplinary teams with context and ownership.
EVIDENCE
Aragon: organization grew from ~7 to ~25; Sophon: led 15 people
SIGNAL DETECTED
0 → 1 PRODUCTS
Build products, test demand and own the business.
EVIDENCE
Fleeber: personally built native iOS/Android products; 100K+ musicians
SIGNAL DETECTED
COST DISCIPLINE
Make technical and product decisions with business economics in view.
EVIDENCE
Sophon: ~$1M annual infrastructure and operating cost reduction
IMPACT ACROSS INDUSTRIES
Consumer
How IOperate
Understand the customer. Choose a direction. Build, measure and adapt.
ACTIVE PATTERN
Understand the customer
Start with the problem people actually have.
I combine interviews, product analytics and competitive research to understand customer needs before choosing a solution. Discovery continues as people use the product.
SIGNALS
- Customer interviews
- Test assumptions
- Usage evidence
- Market context
- Opportunity mapping
EXAMPLES
- Aragon: interviewed customers and competing-product users
- Zenklub: researched profiles, pricing and professional choice
TRACK RECORD
Product decisions. Real outcomes.
The customer problems, choices and products behind the numbers.
EXPLORE ALL EXPERIENCES
Fleeber
MUSIC · SOCIAL PLATFORM · SAAS
2014 — 2018: Founder & CEO
Founded and built a music-tech marketplace and studio SaaS, owning product, technology, UX, GTM, partnerships, sales, growth and operations from concept to native mobile launch and scale.
TRACK RECORD
100K+
Registered musicians
~40K
Peak MAU
~$35K
Studio SaaS ARR
PRODUCT DECISIONS
- Personally built and shipped native iOS/Android products; grew to 100K+ musicians, ~40K MAU and 20K venues and services.
- Personally closed partnerships with EM&T and leading São Paulo recording studios.
- Grew studio SaaS to ~$35K ARR; stopped pursuing scale after discovery exposed structural adoption barriers.
ZKsync
BLOCKCHAIN · INFRASTRUCTURE · ZK
2022 — 2024: Head of Product
Led Product, Design and DevRel across protocol infrastructure, developer tooling and user-facing products; took on overall product leadership following the CPO’s departure.
TRACK RECORD
10+
Chains adopting ZK Stack
$577M
Ecosystem TVL
23M+
Monthly transactions
PRODUCT DECISIONS
- Personally built the first chain-deployment CLI to unblock customer testing and feedback; it evolved into production tooling.
- Led product strategy for ZK Stack, subsequently adopted by 10+ blockchain networks.
- Drove EVM compatibility, account abstraction, smart accounts, wallet SDKs, custody/recovery, portals and documentation.
Sophon
CONSUMER CRYPTO · WEB3 · AI SYSTEMS
2024 — 2026: Chief Product & Technology Officer
Led a 15-person Product, Engineering, Design, Data and DevRel organization while personally prototyping and coding consumer apps, airdrop and claim products.
TRACK RECORD
2
Store-approved apps
~10 weeks
To release-ready each
~$1M
Annual savings
PRODUCT DECISIONS
- Built two release-ready, App Store/Play Store-approved iOS/Android apps: ~6 weeks to V0 and ~10 to release-ready each.
- Validated real purchase flows in production-connected betas across self-custody, bank funding, KYC, cards and DeFi/yield.
- Built and launched 10+ products, integrated 30+ partners and cut annual costs by ~$1M.
PERSONAL R&D
Side Quests
Hands-on AI systems, agent infrastructure and physical interfaces—alongside experiments that taught me when to change direction.
MULTIMODAL / AMBIENT AI SYSTEM
LiveDigital Brain
An end-to-end multimodal AI system I built combining persistent memory, vision, voice, location and tool use across web, mobile, smart glasses and custom robots. Runs across local and cloud models, with real-time scene understanding, memory retrieval, sensor integration and bidirectional hardware control. The product connects intelligent software with the physical world.
Multimodal AIAgentsMemory / RAGSmart glassesHardware control
OPEN-SOURCE AGENT RUNTIME
LiveSamwise OS
An open-source runtime I’m building for modular AI agents: dynamically discovered Markdown skills, hybrid BM25/vector retrieval, structured planning, output validation and repair, and provider-independent local/cloud models. An extensible action layer connects agents to tasks, calendars, storage, search and follow-ups; behavior lives in a separate skills repository.
Agent architectureHybrid retrievalLocal / cloud LLMs
HARDWARE PROTOTYPE
In progressPersonal Robot
A physical interface for Digital Brain that connects custom robots, sensor telemetry, camera and audio with AI reasoning and bidirectional hardware control. The prototype brings together embedded electronics, a screen face, MQTT communication and enclosure design. It is an ongoing experiment in how an assistant can perceive and act in the physical world.
RoboticsESP32MQTTLive Streaming
Web3 Calendly
DeadMeet With
A privacy-first scheduling tool for individuals and groups. The core idea was to provide a solution as good as Calendly without exposing any private data. Users could connect their personal calendars without sharing personal information with whoever was scheduling—or even with us—because all data was encrypted on the client side. Although the idea was well received and the product gained some traction, it never grew beyond early adopters, showing that some core assumptions were wrong. The lessons were: 1. People say they care about privacy, but they often do not act on that concern. 2. Without one strong core value that draws people in, adding multiple good features—even some that are better than competitors'—does not matter. 3. Distribution is everything, and earning initial attention is difficult.
PrivacyCrypto/Encryption
GAME CONCEPT
DeadAmong Claws
An AI-native social deduction game where autonomous agents strategize, deceive, accuse, and compete in real time while humans spectate the drama. The experiment treated agent behavior as the product: a mix of simulation, game design, and multi-agent orchestration meant to generate unscripted humor, conflict, and story. The lesson was blunt: people did not find watching agents play compelling enough, and getting meaningful attention for a strange new format is brutally hard.
GamesMulti-agent AISimulation
THE NEXT PRODUCT CHALLENGE
Build something worth adopting
I’m interested in Head/VP of Product and CPO/CPTO roles where customer discovery, product strategy and hands-on building matter. If you have a complex product challenge and a business to grow, let’s talk.
Testing the next assumption…Understanding the customer…
Looking for a product and technology leader who still builds?
GET IN TOUCHHANDS-ON PRODUCT & TECHNOLOGY LEADER
Which lens you want to use to get to know Ramon?
You can switch later
