edwardpoulin-pepin

Quebec, Canada · local time

Edward Poulin-Pepin

AI-First Product & Engineering Operator

Entrepreneurial full-stack engineer focused on helping organizations structure AI use: turning ambiguous business workflows into specs, adoption plans, AI-assisted development practices, and production systems.

Email me LinkedIn

About

I help teams turn AI from scattered experiments into usable operating systems.

01 AI operating models

Define how teams identify workflows, specify problems, choose tools, and measure AI-assisted work.

02 Specs-driven delivery

Turn business needs into structured specs that humans and AI coding agents can execute safely.

03 Adoption and enablement

Evangelize practical use cases, teach repeatable workflows, and create trust through visible outcomes.

My work experience spans security SaaS, SOAR/SIEM infrastructure, cloud deployment, and full-stack product engineering. My founder projects extend that into Bittensor infrastructure, staking systems, encrypted AI, and AI-assisted product development.

My development style is AI-structured and spec-driven: define the customer problem, turn it into clear product and technical specifications, then use AI-assisted workflows to accelerate implementation while preserving architectural ownership, review discipline, and production judgment.

I care about building systems that create visible user and business outcomes, not just interesting technology.

Work Experience

Professional engineering roles

Experience leading technical teams, translating operational pain into product requirements, and shipping workflow automation in environments where reliability matters.

mondata

Tech Lead / Software Engineer · Security SaaS and infrastructure

  • Lead engineer for a SOAR platform that manages alerts and incidents from multiple SIEMs, including mondata's own.
  • Reduced SOC analyst containment and remediation time by 80% through iterative workflow and product improvements.
  • Managed a remote team of 4 developers, including 2 senior developers, and helped implement ISO-27001 procedures.
  • Designed and implemented incident automation workflows that played a critical role in helping secure $17M CAD in funding.
  • Turned SOC workflows and analyst pain points into clear technical specifications for automation, triage, and incident response features.
  • Built backend and infrastructure systems including automated API deployments on AWS EKS with Terraform, a mono-repo, CI/CD, and automated tests.
  • Earlier, helped build a SaaS SIEM ingesting and converting terabytes of logs daily, and developed a method that accelerated custom EDR file analysis by 200x.

Smartrek Technologies

Developer · Embedded, web, and mobile systems

  • Worked across microcontrollers, web applications, mobile applications, full-stack development, and quality control tooling.
  • Contributed to a resilient, self-healing meshed RF low-power wide-area network.

Venture Work

Founder-led products and technical systems

Part-time venture work where I tested emerging markets, built products quickly, designed operating models, and learned how technical systems create trust with users and stakeholders.

TrustedStake

Founding Engineer · Non-custodial Bittensor staking platform

trustedstake.ai
  • Helped grow TrustedStake, $10M USD in AUM, close to $1M USD in ARR, and roughly $1M USD in daily turnover / rebalance volume.
  • Owned the full-stack platform architecture for a non-custodial Bittensor staking and strategy product, spanning frontend flows, backend services, wallet integrations, validator analytics, rebalancing systems, infrastructure, and release automation.
  • Designed and implemented CI/CD and cloud infrastructure, primarily using Terraform, to support reliable deployments and repeatable operational environments.
  • Helped design a proxy-based staking architecture using Bittensor pure proxies, controller keys, and permissioned transaction flows so users can delegate into strategies without giving up custody.
  • Implemented and optimized the rebalancing engine used to move user stake across Bittensor subnets, reducing execution time from roughly 15-25 minutes to 3-5 minutes.
  • Integrated safety mechanisms including slippage guards, TWAP checks, MEV-shielded execution, and transaction batching logic.
  • Used Specs-Driven Design to translate staking, custody, and strategy requirements into implementation-ready flows for AI-assisted development and human review.
  • Contributed to strategy products, validator economics, cybersecurity planning, audit preparation, onboarding flows, wallet connection tracking, and delegation funnel instrumentation.
  • Worked across TypeScript, React, backend services, Terraform, CI/CD, AWS Lambda, Bittensor/Substrate tooling, signing infrastructure, and blockchain transaction execution.

Miners' Union Validator

Co-Founder · Bittensor validator and decentralized emission voting

minersunion.ai
  • Co-founded and operated a Bittensor validator that reached approximately $100M USD in delegated stake within 3 months.
  • Introduced sTAO, an early delegator-driven subnet emission voting system that gave miners and delegators direct influence over where delegation-related emissions were directed.
  • Helped shift subnet emission voting away from opaque off-chain arrangements controlled by a small number of large validators toward a more transparent and decentralized model.
  • Achieved ecosystem adoption, with competing validators later implementing similar delegator-voting mechanisms.
  • Helped push approximately 30% of total Bittensor delegation toward decentralized subnet emission voting before the model was superseded by the dTAO chain upgrade.
  • Received approximately 113,000 TAO in delegation from the Opentensor Foundation to support stable validation and continued ecosystem tooling.
  • Built validator tooling and analytics for subnet performance, emissions, delegator behavior, stake distribution, rewards, validator transparency, and community alignment.
  • Structured product requirements around delegator incentives, validator transparency, and subnet governance so technical work stayed tied to ecosystem-level outcomes.

Brainlock

Founder / Technical Lead · Encrypted AI and FHE LLM research

  • Founded an encrypted AI initiative exploring Fully Homomorphic Encryption for privacy-preserving machine learning and LLM inference.
  • Designed a product vision where models can operate on encrypted user or enterprise data without exposing the underlying information.
  • Ran a PoC of LLM inference over fully homomorphically encrypted data, proving the privacy-preserving execution path was possible, although still too slow for practical product use.
  • Evaluated Zama Concrete-ML and Opentrust Labs technology, then explored whether Opentrust Labs could be optimized for matrix and tensor operations required by private LLM execution.
  • Developed early architecture around cryptographic bindings, FHE context management, ciphertext operations, encrypted inference constraints, and a Bittensor subnet concept for private model execution.
  • Paused the project after determining the underlying FHE tooling was not yet mature enough to support optimized on-chain private LLM workloads.
  • Used AI-first product exploration to move from privacy problem framing to concrete architecture, technical constraints, PoC implementation, and partner evaluation.
  • Explored commercial licensing and partnership paths with FHE providers.

9516-7284 Québec Inc.

Founder / Operator · Bittensor mining and crypto infrastructure

  • Founded and operated a Canadian corporation focused on crypto infrastructure, initially centered on Bittensor mining and subnet participation.
  • Built and operated mining infrastructure across multiple subnets, optimizing performance, uptime, subnet selection, registrations, and reward capture.
  • Developed operational tooling to monitor subnet performance, miner health, emissions, validator behavior, reward dynamics, and registrations.
  • Built operational specs and feedback loops for miner performance, uptime, subnet selection, and reward capture.
  • Managed corporate-level crypto treasury exposure, including mined TAO accumulation and trading.
  • Built hands-on production experience with Bittensor infrastructure, Substrate tooling, signing flows, wallet infrastructure, hotkey/coldkey management, and blockchain operations.

Tripweave - Work in Progress

Founder / Builder · Round-the-world trip planning MVP

  • Built a round-the-world trip planning MVP focused on complex multi-country itinerary planning.
  • Designed workflows for itinerary generation, destination sequencing, trip organization, destination discovery, and constraint-heavy planning.
  • Built the frontend with Vite, TypeScript, React, shadcn-ui, and Tailwind CSS.
  • Applied a fast, specs-driven MVP approach: define the customer pain, turn it into a usable workflow, build quickly, and validate from there.
  • Minimally Viable Product to be released soon.

Tool Stack

Capabilities for structuring AI adoption and shipping the systems behind it.

A mix of AI enablement, product architecture, full-stack engineering, infrastructure, and emerging crypto/AI systems experience.

AI Enablement

AI-first development AI-structured workflows AI agent direction Adoption playbooks Workflow mapping

Product & Specs

Specs-Driven Design Requirements definition Product architecture Product strategy Go-to-market thinking

Full-Stack & Infra

TypeScript React Python Django Angular Backend services Frontend systems Terraform Docker Kubernetes AWS GCP CI/CD architecture PostgreSQL Elasticsearch

Crypto & Emerging Systems

Bittensor Substrate Validator infrastructure Non-custodial staking Wallet integrations Blockchain signing flows MEV-shielded execution FHE / encrypted AI Shadeform VAST

Education

Université Laval

Pairing technical execution with customer discovery, positioning, segmentation, and go-to-market thinking.

Market layer Building the bridge between what customers need and what AI-enabled teams should build.

Marketing Certificate

Completing a marketing certificate to complement software engineering with customer discovery, product positioning, segmentation, and go-to-market strategy, with the goal of integrating market insight from the earliest product and architecture decisions.

Bachelor of Computer Engineering

Completed in 3 years, ahead of the conventional 4-year timeline, while balancing full-time internships and jobs.

Available for AI enablement and product engineering roles

Building AI adoption into real workflows?

edward.pepin@outlook.com Telegram @edwardpou LinkedIn profile