Signals in.
Revenue out.
I build the machine.

Edgar Alvarado, GTM engineer. I design programmatic outbound engines, signal-based prospecting, and revenue infrastructure for B2B teams in AI and cybersecurity. Below is the kind of system I build, running live on this page.

Scroll to see it run ↓
the_engine // live simulation source → enrich → score → route
every dot is an account · most never deserve an email qualified: 0
stage 01 // shipped systems

Four machines. The last one is hiring me.

The first three run confidential: the architecture patterns are mine to show; the numbers and blueprints belong to the companies. The fourth is fully open: I own it, and it’s running right now.

module_01 // outbound_engine ● shipped & operating

The Programmatic Outbound Engine

A manual prospecting motion couldn’t cover the addressable market. I architected an automated revenue engine: enrichment and orchestration, workflow automation, and AI-driven personalization at scale, feeding clean, high-intent records into the CRM, with self-serve dashboards tracking full-funnel conversion.

It produced a meaningful, sustained pipeline contribution, generated by systems rather than added headcount.

context // under nda
replaced Manual prospecting, list-buying, copy-paste personalization
stack
Clayn8nAI agentsSalesforceLooker
module_02 // greenfield_architecture ● delivered as blueprint

Greenfield GTM Systems Architecture

A complete outbound architecture, designed from a blank page: signal-driven sourcing that qualifies accounts before a dollar of enrichment is spent, isolated sending infrastructure that protects the primary domain, and a deliberately clean CRM where only genuine hand-raisers ever become records.

Attribution is wired through the whole chain. Every meeting booked reports back to its original signal source, so the system continuously optimizes its own spend.

context // under nda
design goal Scale pipeline without scaling headcount, built for a lean GTM team
stack
ClayCold email infraHubSpotWebhooks & APIs
module_03 // signal_intelligence ● methodology, reusable

Signal Engineering & ICP Intelligence

Weighted scoring models that identify a market’s highest-intent segment from publicly observable signals: technographics, hiring patterns, and developer activity. Outreach only ever reaches accounts already leaning toward the category.

Accounts below the threshold never consume a dollar of enrichment or a single send. The market filters itself before outreach begins.

context Independent research · emerging-tech markets
principle Precision beats volume. Spend follows intent, never the other way around.
stack
ClayIntent dataPublic APIsSales Nav
module_04 // open_blueprint ● running right now

The Job-Search Engine

The one machine I can show you completely, because the client is me. My job search runs as a GTM system: target companies are accounts, open roles are deals in a CRM, and a versioned fit-score rubric (hard gates, fourteen weighted signals, an evidence floor) decides where outreach effort goes. Every score ships with a confidence rating; thin evidence reads as “insufficient data,” never as a number.

Scheduled AI agents sweep the pipeline three mornings a week, verifying roles are still live, rescoring on new evidence, flagging stale deals. The rubric is on v3 because v1 embarrassed itself: it let a company score 78 while failing a non-negotiable requirement. Weights flatter; gates decide. That lesson is now architecture.

context My own job search. No NDA. Ask for the live demo.
the point If I build this to find one job, imagine what I’ll build for your pipeline.
stack
HubSpotClayMakeClaude agentsScheduled automations
stage 02 // operating disciplines

Four disciplines, one operating system.

Signal in, revenue out. Nothing manual in between that a system could do better.

01 / detect

Signal Engineering

Weighted scoring on real buying signals: hiring intent, tech stack, developer activity. Outreach only targets accounts already leaning in.

02 / deliver

Outbound Infrastructure

Isolated sending domains, automated rotation, deliverability defense. Volume that scales without ever risking the primary domain.

03 / measure

CRM & Revenue Observability

CRM architecture where every record earns its place. Attribution wired through, so cost-per-qualified-lead is a report, not a guess.

04 / compound

AI-Assisted Workflows

Agentic enrichment, LLM-driven personalization, automated research loops. AI where it compounds, deterministic logic where it counts.

stage 03 // signals.log

Field notes from the build.

Short, opinionated notes on GTM systems, written when something actually ships, not on a content calendar. Every claim verified against the changelog, not the hype thread.

2026-08-03 · scoring

Your scoring model is wrong until it embarrasses you

v1 of my fit-score rubric let a company score 78 while failing a non-negotiable requirement. The fix wasn’t better weights. It was admitting that some signals aren’t signals, they’re gates. If a criterion is non-negotiable, it doesn’t belong in the average. Weights flatter; gates decide.

2026-07-25 · agents

Account research just became a standing process

Clay’s research agents reason over your combined call, CRM, and email data to keep account intelligence current. That used to be an hour of manual stitching per account. The shift that matters: nobody has to update the CRM for the agent to stay current. Research stops being a task and becomes infrastructure.

2026-07-25 · spend control

Hard spend caps are a GTM feature, not a finance feature

Every enrichment loop should carry its own ceiling. Give each pipeline its own API project with a hard cap, alert at 70%, and catch the cap error distinctly instead of retrying. This kills the “the loop ran all weekend and burned the month’s budget” failure mode: the quiet tax on every automated GTM motion.

stage 04 // the operator

From carrying the quota to building the machine.

I spent eight-plus years inside cybersecurity go-to-market: enterprise SDR, account executive, revenue operations. I’ve carried the number, missed the number, and watched brilliant sellers burn out doing work a system should have done for them.

That’s what pulled me into GTM engineering: the conviction that pipeline is an infrastructure problem. The best revenue teams don’t out-hustle the market. They out-architect it.

Today I build that infrastructure: enrichment waterfalls, signal-based routing, AI-assisted personalization, and the reporting layer that proves what’s working. I’ve run the same architecture on both Salesforce and HubSpot, just different nomenclature for the same underlying pattern. Based in the San Francisco Bay Area.

career.log // trajectory
▸GTM Engineer
▸Growth Marketing Operations
▸Revenue Operations
▸Account Executive
▸Enterprise SDR
stage 05 // routing

You made it through the funnel.

Open to GTM engineering and revenue systems roles, and always glad to compare notes on outbound architecture.

route → edgar@ealvarado.com linkedin