Currently accepting new engagements

Anthony Paquet

I help engineering leaders ship faster, cut costs, become secure and compliant.

Builder & Engineering Leader, results in weeks, not quarters. No long-term contracts required.

AI & Cloud ExpertSecurity and complianceFinOpsFortune 500 & Startup Experience
Explore my personal projects
Anthony Paquet
AI & Cloud Architect | Engineering LeaderQuebec, Canada

Anthony Paquet

AI & Cloud Architect | Engineering Leader

I design and scale secure, cost-efficient cloud systems and take products from 0→1 to enterprise scale. Deep FinOps work cuts cloud spend, and a DevSecOps approach helps teams ship faster without sacrificing security. The result is resilient infrastructure that holds up under real load and keeps costs down.

Knowledge is the ultimate wealth. The more you share, the more it grows. I believe in continuous learning and empowering others through expertise.
Proud dad of twoAvid snowboarderTech passionate
Quebec, CanadaFrench & English

Agent in the loop.

I talk to AI from my phone. It turns my intent into a plan, keeps the useful context with the task, then builds and verifies the result on its own. I remove blockers so every loop is as short as possible, from local implementation to customer feedback. The same approach connects engineering, infrastructure, and go-to-market.

Sitting down to type a prompt is already one step too many.
The targetBusiness outcomes.shortest loop
  1. 01Speak. I give AI the intent, context, and outcome I want, by voice.
  2. 02Plan. AI turns the request into a concrete plan, tasks, acceptance criteria, and expected evidence.
  3. 03Unblock. Access, data, environments, tests, and previews are ready so it can act without waiting.
  4. 04Design. The business domain leads. Ports and adapters keep tools, models, and clouds replaceable.
  5. 05Build. AI builds the first useful slice and opens the sub-loops it needs.
  6. 06Iterate. It works locally against the plan: run, observe, fix, and repeat, with context attached to the task.
  7. 07Test. Each domain is checked with unit tests, BDD, and controls matched to the risk.
  8. 08Review. A second scheduled task, on another model, cross-validates the result and flags gaps before release.
  9. 09Ship. Depending on the use case, we ship to one customer, a beta cohort, or everyone. Versions can coexist when needed.
  10. 10Learn. Real usage and customer feedback show what works, what blocks, and what creates value.
  11. 11Improve. The agent updates the product, the plan, and the offer from evidence, not assumptions.
  12. 12Repeat. We start again with context, learning, and access already in place, making the next loop shorter.

Consulting · engineering leadership.

Flexible engagement models with clear outcomes. From 0→1 to enterprise scale. Results in weeks, not quarters.

AI Infrastructure • LLM Security • Agents & Evals

Take AI from demo to dependable. Infrastructure, security, evals, and agent engineering for teams shipping real AI products.

AI Infrastructure & Eval Platforms

Production-grade infrastructure for AI agents and model evaluation. Isolated sandboxes, GPU cost control, and the platform engineering behind reliable AI products.

AI / LLM Security & Governance

Ship AI to production safely with LLM gateways, PII protection, prompt-injection controls, and the audit trails that pass SOC 2.

AI Adoption & Agent Engineering

Turn AI from a demo into a dependable system with agent design, evals, and tooling that move real business outcomes.

Experience

Proven outcomes across cloud, security, and infrastructure.

Founder & AI Orchestrator

Nota

Quebec, Canada

Nota is a marketplace for notarial acts across Québec. Clients post the act they need, the date they want to sign and the price they offer. Every request appears on a public calendar, and notaries take the ones that fit their schedule. Posting a date is free.

One AI agent per departmentMarketing, go-to-market, sales and engineering
Solo operatorEvery function, orchestrated with AI
  • Founded Nota to modernize the Quebec notary industry, an underserved, paper-heavy space overdue for software built around how notaries actually work.
  • Run every function of the company as its sole operator, orchestrating AI agents built on the latest models from Anthropic (Claude), OpenAI and xAI.
  • Each department has its own AI agent for marketing, go-to-market, sales and engineering, coordinated through my orchestration.
Legal TechNotary TechQuebecFounderAI Agents0→1

Certifications · education.

AWS Certifications

AWS Certified Solutions Architect, Associate

AWS Certified Solutions Architect, Associate

Amazon Web Services (AWS)

AWS Certified Developer, Associate

AWS Certified Developer, Associate

Amazon Web Services (AWS)

AWS Certified Cloud Practitioner

AWS Certified Cloud Practitioner

Amazon Web Services (AWS)

Education

Computer Science

Cégep Garneau

2008 to 2011

Québec, Canada

Languages

FrenchNative
EnglishFluent

Ideas I build into products.

Independent projects at the intersection of useful software, human interaction, and entrepreneurship.

A public calendar where clients across Québec post their notarial act, date and price, and notaries pick the requests that fit.