I build the AI tools and systems email teams run on.

Agents, LLM workflows, internal tools and lifecycle platforms, with the QA and governance to run them at scale. Braze, HubSpot, Iterable, Salesforce Marketing Cloud and more.

Companies

Teams I've built for

Companies10
AnthropicAnthropic
Apple
Google Fiber
DoorDash
Indeed
Target
OpenSea
Strava
InboxMonster
FITURE
Platforms6
Braze
HubSpot
Iterable
Salesforce MC
Customer.io
Zeta

Skills

What I bring to a team

AI agents and LLM workflows

Agents, prompt libraries and API integrations that automate campaign operations, content and reporting.

Structured promptingTool useAPIs

Internal tools

Self-service builders and apps that let marketers ship without waiting on engineering.

ReactNext.jsTypeScript

CRM platform ownership

Architecture, administration, roadmap and reliability for the platforms lifecycle runs on.

BrazeHubSpotIterableSFMC

Leadership

Led developer teams, set standards and partnered with Marketing, Data, Product and Engineering.

Team leadMentoringCross-functional

Data, segmentation and personalization

Marketing data requirements, integrations and personalization frameworks across channels.

LiquidAMPscriptHubLHandlebars

QA and governance

QA standards, automated testing, link validation and guardrails before every send.

Automated QAGovernanceGuardrails

Workflows

How I automate email operations

AI where it helps and plain automation where it's the better tool, with guardrails, a person in the loop, and numbers that show it's working.

  1. 01

    Intake

    Requests from forms, tickets or Slack become structured specs.

  2. 02

    Templates and prompts

    Reusable templates, rules and tested prompts for each task and brand voice.

  3. 03

    Automate

    Workflows and agents that call your systems: CRM, ESP, assets and analytics APIs.

  4. 04

    Guardrails

    Deterministic checks run on every output before anyone sees it.

  5. 05

    Review and approve

    Marketers approve, edit or reject in one place.

  6. 06

    Measure

    Adoption, time saved and error rates, so the value is visible.

Inbox QA

Paste an email. Get a report.

The checks that break email in real inboxes, with one click fixes for the safe ones. Try the sample, or paste your own HTML.

fall-promo.html
email.html Ln 1
0

Checking

    Runs in your browser

    What this shows: automated testing and QA standards, the kind of gate I set up so lifecycle teams catch problems before a send. Built from scratch.

    Dynamic email

    Your data, in every ESP's language

    Customer data decides what each subscriber sees. AI picks the segments and drafts the variants, and the template logic gets the right one to every person, on any platform.

    Pick a subscriber

    Profile data

    
            

    What lands in the inbox

    Northwind Coffee
    fall-promo / hero module
    
          

    What this shows: segmentation and personalization logic written natively for each platform, from Braze Liquid to HubSpot HubL.

    Selected work

    Email I've built

    Apple email: Deck it out, accessories for iPhone 17 Pro
    AppleiPhone 17 Pro
    DoorDash email: 20% off pet supplies for National Dog Week
    DoorDashNational Dog Week
    OpenSea Currents newsletter, volume 20
    OpenSeaCurrents newsletter
    GoPro email launching HERO9 Black
    GoProHERO9 Black launch
    Google Fiber offer email
    Google FiberUpgrade offer
    OpenSea creator interview email
    OpenSeaCreator interview
    GoPro Black Friday email
    GoProBlack Friday
    InboxMonster welcome email
    InboxMonsterWelcome series

    Contact

    Let's talk.

    Hiring for an AI email or lifecycle platform role, or need help with one? My inbox is open.

    Send me an email
    webdevcastillo@gmail.com