Selected work across growth, revenue operations, positioning, AI-assisted systems, and internal builds. Prior in-house roles, client engagements, and self-commissioned projects are labeled separately so the context stays attached to the result.
The through-line is simple: understand the commercial problem, build the system underneath it, and do not claim more than the evidence supports.
A growth engine built across marketing, RevOps, SalesOps, CRM, and cloud marketplaces.
Labra builds the platform software companies use to sell through AWS, Microsoft, and Google cloud marketplaces.
I owned the operating layer behind growth: enterprise demand generation, RevOps, SalesOps, CRM architecture, reporting, and the cloud-marketplace motion.
The result was more than $6M in net-new qualified pipeline built from near zero, 182% year-over-year revenue growth, quarter-over-quarter coverage above 3×, and a 32% reduction in deal-cycle length.
The systems work mattered as much as the campaigns. I re-architected the CRM and built structured data and automation around the sales process.
The clearest antecedent of the work Dandelion Fire does now. Fastbot was a GPT-based assistant embedded directly into Slack and CRM workflows: structured outputs, instruction hierarchies, compliance guardrails, and human review on customer-facing steps.
It ran inside a live revenue operation rather than a pilot, which is where the real constraints show up. The model was never the hard part. The hard parts were giving it enough company context to be useful, deciding which steps a person had to approve, handling the cases nobody specified, and getting a team to trust it enough to keep using it.
Built in-house, in a prior role. Not a client deployment, and not evidence of what an engagement produced. It is the reason the Implementation Sprint is scoped the way it is, with context, exceptions, evaluation, and a named owner treated as part of the build rather than afterthoughts.
Growth strategy and revenue architecture on the same desk.
Shelf builds AI-powered knowledge infrastructure for support and contact-center teams.
I led growth marketing and revenue operations simultaneously, directing a team of four while rebuilding the martech stack, demand engine, reporting, and the revenue architecture underneath them.
During that period, the company surpassed $10M ARR with back-to-back $1M+ new-logo quarters. Pipeline grew 3×, and AI-powered dynamic landing-page work increased demo-booking conversion 4×.
Holding both roles at once shaped Dandelion Fire's operating view: marketing strategy, data, process, and implementation are not separate problems when the buyer experiences them as one system.
Repositioning an AI systems company for regulated industries, then building the GTM system around the new story.
Botable builds custom AI systems for regulated, knowledge-dense operations in MedTech, pharma, and life sciences, then stays to operate them.
The initial problem was commercial: the market could read the company as another SaaS chatbot, while the actual delivery model and regulatory context were much more substantive.
The work has included positioning, narrative, brand system, website, sales story, partner enablement, and the outbound operating model built around them.
The engagement also became a test bed for structured AI-assisted GTM work: research, account qualification, claim review, content production, and human approval designed as repeatable processes rather than isolated prompts.
Scaling content operations while building the intelligence system underneath them.
Project lead for corporate social on behalf of a major enterprise-software investor, contracted through the digital marketing agency that serves as the firm's agency of record.
The engagement combines content operations, measurement, and competitive intelligence. I built a competitive audit across 11 PE and VC firms, an executive-content performance study from first-party data, and an agent-driven competitive-intelligence workflow with human review on client-facing output.
Over the measured six-month comparison period, publishing volume nearly tripled while total engagement grew 9.6% and net audience growth accelerated 26.2%.
Source: platform analytics comparing the six months before engagement with the engagement period.This is client work involving AI-assisted operations. It is not a full client AI-transformation deployment.
The system I use to learn what agentic work requires in practice.
Atelier is Dandelion Fire's internal agentic workbench. It exists to manage context, work orders, review, research, and the operating chain behind AI-assisted work. The system is continuously evolving as real work exposes what agents need, where they fail, and where human judgment belongs.
It is evidence that Dandelion Fire builds and operates agentic systems internally. The principles that matter: structured context instead of isolated prompts, defined agent jobs, explicit review and approval, evidence-linked outputs, logging and traceability, and ownership and handoff.
It is not evidence that this exact system has been deployed for a client.
A small world built because the technology should still feel surprising.
The Field is a first-person interactive meadow, puzzle, and second world scored with my own music. It is built in-house with AI assistance and runs directly in the browser, and it sits at the top of the homepage.
It proves something narrower and more useful than a client result: I build with the technology I sell, and I care what the result feels like.
The Field is not a client engagement, a GTM result, or proof that a company should commission a WebGL world.
The goal is not to keep people manually involved in every step. The goal is to know which work an agent can own, which decisions require judgment, and how the system proves it did what it was supposed to do.
The system gets the business context required to do the job: criteria, positioning, source material, data, constraints, and examples.
Every agent or automation has a job, an expected output, and boundaries.
Output is checked against the requested work and the evidence available.
High-judgment and external actions pass through the agreed approval point.
Important decisions, source material, exceptions, and changes remain inspectable.
The system proposes. A person decides.
This is not governance theater. It is how an AI capability becomes something a team can rely on.
Bring the initiative, workflow, or recurring work. We will talk through what a useful implementation would need to do and whether Dandelion Fire is the right partner to build it.
Discuss your AI GTM priorities →Prefer to send context first?