Selected work

Problems I have solved

Three ways of doing the work: owner of the operation, external consultant, and operator inside a team.

Operator inside a team

Axenya · B2B, Corporate health benefits

01

From a five-day forecast ritual to real time

4 days → real time 8 panels · 20 releases 1,000+ properties audited
The challenge

Axenya, a B2B corporate health broker with 81 people, ran forecast by exporting the entire pipeline to Excel and recalculating it cell by cell, by hand. Four to five days of work, repeated every two weeks, for a pipeline in the hundreds of millions in ARR. Nobody on the product team had ever looked at it.

My role

Revenue Operations Analyst, four months. I translated revenue leadership’s strategy into a working system, then built that system myself: architecture, the dashboards, the rules, the rollout.

What I built
  • Shipped a dashboard platform from zero: 8 panels covering CRO, Forecast, Board, AE, BDR, Last 48h, Delta, and Coverage views, each chart auditable back to the exact CRM fields behind it.
  • Built a semantic catalog as the single source of truth for every business rule, with documentation generated straight from it. Building it surfaced two probability rules running side by side unnoticed: the same stage was worth 18% on one screen and 33% on another.
  • Turned a meeting into a written four-step forecast ritual: per-rep prework, named data-hygiene follow-up, post-meeting to-dos, and a consolidated report showing the delta against the previous cycle.
  • Rebuilt the CRM’s business-rule layer as fields and automations, not verbal agreements: ARR modeling by compensation type, automatic company-size classification, a canonical close-date definition, and automated inactivity handling.
  • Shipped a daily Slack pipeline report and automated field-hygiene nudges per rep, each with a direct link to the record.
  • Remapped handoffs across Sales, Quoting, and Implementation: correlated fields, shared quote files, and stalled-quote alerts, surfacing roughly 60 stuck deals during rollout, some tied to reps who had already left.
The human part

The infrastructure never went down, not once in three months in production. Every hour of rework came from a business premise that changed mid-cycle: does a POC count as revenue, is stage four active pipeline. The machine was never the bottleneck. The person deciding was. I found the same pattern in adoption: reps were not filling fields because the screen was full of information that mattered to RevOps and to nobody selling. There were fields that mattered to me and to nobody else. So I stopped asking for them upfront.

HubSpotVercelZapierSlack APIClaude CodeGitHub

Owner of the operation

Nola · B2B SaaS, Restaurant management

02

Building the revenue engine from zero

SDR 39% → 95% CAC −76.8% Revenue +569%
The challenge

Nola, a B2B SaaS for restaurant management, had no revenue infrastructure. No CRM architecture, no pipeline visibility, no automation. Ten SDRs working by hand, reaching fewer than 4 in 10 leads.

My role

I led the marketing and revenue operations function, managing a team of three (paid media, social, and design), building it from scratch alongside the Head of Sales.

What I built
  • Architected the HubSpot ecosystem: pipeline, custom properties, lead routing, permissions, and data hygiene.
  • Engineered a lead-connection automation (HubSpot + n8n) that raised SDR connection from 39% with 10 SDRs to 95% with 3.
  • Unified paid media and inbound into one attribution model via RD Station and Zapier.
  • Scaled demand generation with predictive lead scoring and full-funnel MQL/SQL definitions.
The human part

The connection automation worked because it fit how the SDRs already worked. The team adopted it, and adoption is what made the numbers real.

HubSpotn8nRD StationZapier

External consultant

B2B SaaS · Construction tech, via Deal Machine

03

A full HubSpot implementation, from scratch

83 LPs standardized Full attribution stack Sales + CS pipelines live
The challenge

A construction-tech SaaS running active campaigns across Google, Meta and LinkedIn, with no CRM structure, no conversion tracking, and 83 landing pages capturing data in incompatible formats. Leads arrived. The pipeline was invisible.

My role

Engaged through Deal Machine to design and implement the entire CRM foundation under a defined project scope.

What I built
  • Implemented HubSpot end to end: CRM architecture, object mapping, Sales and CS pipelines.
  • Built custom fields for ICP qualification and UTM tracking for end-to-end attribution.
  • Standardized lead capture across 83 landing pages via GTM: pixels, events, and form normalization.
  • Integrated the full ad stack into HubSpot and designed the lead scoring and nurture framework.
The human part

Delivered as an external partner: in, diagnose, implement, hand over. A system the internal team could run without me.

HubSpotGTMGoogle AdsMeta AdsLinkedIn Ads

Growth Manager, two sister brands

Arcca Holding · Nola (SaaS) + Grupo Arcca (Franchising)

04

Running growth for two sister companies at once

Nola: lead base +1,184% Arcca: CPL −48%, ROAS <90 days MQLs +220% · SQLs +182%
The challenge

Nola (B2B SaaS) and Grupo Arcca (a franchise portfolio) are sister companies under Arcca Holding. Both needed a growth engine built from zero, at the same time, with no shared playbook between them.

My role

Growth Manager at both, concurrently: Marketing Operations & Growth Manager at Nola, Marketing Manager at Arcca. Led the Design, Social Media, and paid media teams across both.

What I built
  • At Nola: scalable inbound funnels with predictive lead scoring, growing the lead database 1,184% in 17 months, MQLs 220%, SQLs 182%.
  • At Arcca: the franchise lead-gen funnel from zero, raising contact rate from 3% to 20-26% and cutting lost leads from 98% to 31-42%, ROAS positive in under 90 days.
  • Directed paid media and digital campaigns across both portfolios, tracking ROI and CPL by channel and brand.
  • Repositioned Arcca Franchising’s brand identity while keeping Nola’s demand-gen and RevOps build moving in parallel.
The human part

Running growth for two companies at once only works if neither team feels like the side project. I gave each its own cadence and its own team, Design and Social Media at Arcca, Marketing Ops at Nola, and made sure both moved forward every week.

HubSpotRD StationGoogle AdsMeta AdsZapier

Self-directed build

Personal brand · Career transition

05

The site you’re reading right now

Bilingual EN/PT from day one Zero frameworks, zero dependencies Schema.org + llms.txt from launch
The challenge

Leaving a role and needing a portfolio that proved the actual thesis I sell: systems built around the people who use them. Without months of runway or a dev team to hand it to.

My role

Product owner, writer, and QA. I directed an AI pair-programmer (Claude Code) for implementation while owning every call on positioning, content, structure, and design myself, in both languages.

What I built
  • Defined the positioning and wrote the voice in EN and PT-BR before a line of code existed.
  • Built a single bilingual system instead of duplicated pages: one dictionary, both languages share the same structure.
  • SEO and answer-engine visibility from day one: schema.org on every page, canonical URLs, sitemap, and an llms.txt file most sites skip entirely.
  • Audited my own links before launch and fixed a routing gap, the same QA discipline I would run on a client's CRM.
  • Applied the same discipline as the Axenya build: documented for handoff, adoption as the metric, not effort.
The human part

The site had exactly one user to please before anyone else saw it: me, reading every sentence and clicking every link like a stranger would. Same discipline as a CRM audit, just pointed at myself.

Claude CodeHostinger VPSSchema.org

A version of this same build also runs as a fixed-scope project, a V0 in two to three weeks, shaped around your sales cycle and revenue model. No package, no pricing table. If it is relevant, the same LinkedIn or email above works.

If this way of seeing the work resonates

No pitch. Just connection. Open to new Revenue Operations roles.