AI systems custom-built by someone who’s run sales, marketing, and ops.Not just advised them.

Fifteen years of enterprise B2B. Built, shipped, still running.

View the workBook a call

The path

Both halves of a deal, in one person

  1. Before 2026

    Fifteen years running enterprise B2B

    Ran sales, marketing, and ops inside enterprise B2B, from the seat, not the slide deck. That's the vantage point behind every system on this page.

  2. 2026 · 01

    Signal-Based Lead Generator→

    Rejected the category: stop renting lead lists, build buying-window radar instead.

  3. 2026 · 02

    LinkedIn Content Engine→

    Rejected the volume play: a slop filter, not a content firehose.

  4. 2026 · 03

    Sales Navigator Outreach System→

    Rejected the tier ladder: built the integration LinkedIn won't sell (inside ToS).

  5. 2026 · 04

    Brand Voice Operator→

    Rejected the document: a voice a new hire could run on day one.

  6. 2026 · 05

    Launch Command Center→

    Rejected the black box: auditable autonomy you can open and inspect.

Featured

Selected work

Four production systems, each rejecting the obvious category.

A market of dim gray dots, a few glowing amber and connected by lines that converge into a single pipeline.
Case studyLead Gen & Outbound

Signal-Based Lead Generator

Detects buying windows from public compliance signals, before the buyer starts shopping.

TrustNet · 2026→
Dense raw source material refined through processing lines into an evenly spaced publishing rhythm of amber blocks.
Case studyContent & Voice

LinkedIn Content Engine

A staged LinkedIn pipeline with two quality gates: one for topics, one for drafts.

Personal · 2026→
A single amber voice-signature waveform reshaping nearby gray text blocks to match its cadence.
Case studyContent & Voice

Brand Voice Operator

Turns a company's own writing into an evidence-backed voice guide a team can run.

Open source · 2026→
Three parallel market timelines with progress dots, amber risk flags, one breached deadline marker, and a review scan line.
Case studyIntelligence & Agents

Launch Command Center

An auditable AI layer over a healthcare launch: it reads, drafts, and flags, never decides.

Confidential client · 2026→

Everything

All work

Case studies and live demos. Filter by focus.

Process

How I work

Every engagement runs the same shape: decide what's worth building, build it in stages, and hand it off so your team can run it. A human gate sits between every phase. Nothing advances without your sign-off.

01

Audit

I map where the hours actually go and which tasks are worth automating at all. You get a written opportunity map before any proposal.

02

Roadmap

A staged plan ranked by payoff and risk, with what “done” means for each stage. You approve the sequence before code exists.

03

Gated build

I build in stages behind approval gates. Drafts are the default; nothing sends, posts, or spends until you open the gate.

04

Operating handoff

Runbooks, docs, and a walkthrough so the system runs without me. I design myself out of the loop on purpose.

Under the hood this runs on ICM, a file-based methodology where every stage has a contract and an audit trail. Ask me about it on a call; it’s the part I’ll talk about too long.

Let’s talk.

Hiring, contracting, or want to think through a build? Pick any of these.

Email meLinkedInGitHubBook a call

raph@raphaelmercado.devlinkedin.com/in/raphmercadogithub.com/raphaelmercado-coder