Measure, don't guess.
Every decision earns its keep with a number. Hunches are for hobbies; work runs on data.
[ 01 ] Engineer · marketer · builder
I build, ship & scale things — then measure if they worked.
Performance marketer turned full-stack engineer. I write code, run campaigns and wire automations — then measure whether any of it actually moved a number. This site is where I keep the principles, the stack and the writing.
- 10+ yrs in the trenches
- $50M+ ad spend run
- Ship most Fridays
1// build.ts — tech that compounds
2import { Builder } from "@asharma/dev";
3
4const project = new Builder({
5 stack: ["TypeScript", "Next.js", "Postgres", "n8n"],
6 focus: "marketing-grade systems",
7 velocity: "2-week sprints",
8});
9
10await project.ship({
11 ci: true,
12 tests: "real",
13 roi: "measurable", // not vanity metrics
14});● build passingtests greendeployed → production
By the numbers
$50M SPEND1,000+ CAMPAIGNS80+ PROJECTS10+ YEARSREACTNODE
// 02 How I think
Principles I build by.
Ship on Fridays.
Tests, rollback, monitoring. Caution is a tax you pay on weak foundations — not a virtue.
The funnel is the product.
Acquisition is a first-class engineering concern, not a marketing afterthought. A leaky funnel is just a bug in production.
Boring tech, hot problems.
Postgres, cron and TypeScript have shipped more revenue than this week's new framework. Reach for the clever thing last.
Compounding beats clever.
Small systems running daily outperform big launches. The dashboard you actually open every morning beats the one you don't.
Automate the repetitive.
If I do it twice, I script it. If it runs weekly, I schedule it. Time is the only thing I can't scale.
// 03 How I work
How I approach the work.
- 01
Discover
Before any code, I pressure-test the idea, map the scope and de-risk the unknowns. Most projects die here — saved.
Start - 02
Architect
I pick the boring stack, model the data once, and stand up CI/CD, staging and observability before the foundation sets.
Foundation - 03
Build
Two-week sprints, working software every Friday. Tests, docs and reviews land on every change — not as a phase.
In motion - 04
Ship
I launch with load tests, rollbacks and monitoring in place. Yes, even on a Friday.
Out the door - 05
Scale
Iterate, measure, harden. Code I write is code anyone can read, run and extend — including future me.
After
$0M+
Ad spend run
0+
Years experience
0+
Campaigns run
0+
Projects shipped
// 04 My toolbox
The stack I reach for.
TypeScriptReactNext.jsNode.jsPythonPostgreSQLMongoDBGraphQLDockerTailwind CSSn8nZapierMakePipedreamGoogle Apps ScriptGA4BigQueryLooker StudioMeta AdsGoogle AdsOpenAIAnthropicLangChainRAGQdrantMCPAI AgentsEmbeddingsWhisperFunction CallingCrewAIOllama
// 05 Latest writing
From the blog.
Data & Analytics
Marketing Mix Modeling Is Back: Build an MMM That Actually Works in the Post-Cookie Era
Paid Media & MarketingGoogle Demand Gen 2026: Display Ads Are Gone—Here's How to Build Full-Funnel Demand That Actually Converts
Data & AnalyticsMarketing Attribution Models in 2026: Why Last-Click Is Dead and What Replaces It
Paid Media & MarketingAgentic Commerce in 2026: When AI Shopping Bots Buy for Users — How Brands Stay Discoverable and Purchasable
Paid Media & MarketingChatGPT Ads Manager in 2026: How Conversational Intent Changes Everything About Where You Spend
AutomationAI Agents for Marketing Automation: Beyond Workflows to Autonomous Campaigns
Paid Media & MarketingReddit Ads in 2026: The Underrated Performance Channel Booming on AI and Commerce
AutomationAI Chatbots for Lead Qualification: How to Build One That Actually Filters Leads
Paid Media & MarketingGoogle AI Max for Search: The DSA Replacement and Your Migration Action Plan
Paid Media & MarketingGA4 Server-Side Tagging in 2026: How to Save Your Ad Attribution Before Cookies and ITP Kill It
AutomationMarketing Automation with n8n: How to Build a Self-Hosted Stack That Replaces $500/mo in SaaS Subscriptions
Tools & ReviewsBuild a Real-Time Marketing Dashboard: Pull 5 Ad Platforms Into One Looker Studio Report
Data & AnalyticsPredictive Lead Scoring With AI: Stop Guessing Which Leads Will Close
Data & AnalyticsAI Email Personalization Beyond Hi First Name: Real 1:1 Personalization at Scale
Paid Media & MarketingIncrementality Testing for Paid Media: Did Your Ad Actually Drive the Sale?
Tools & ReviewsAI Ad Creative at Scale: Generate Hundreds of Variations With Midjourney, DALL-E, and Runway
Paid Media & MarketingMeta Advantage+ Campaigns in 2026: When Full AI Automation Works (and When It Burns Money)
Paid Media & MarketingGoogle Performance Max: How to Actually Control the AI Black Box (Asset Groups, Audience Signals, Budget Caps)
Tools & ReviewsBuild vs Buy vs Automate: A Decision Framework for 2026
Paid Media & MarketingMeta Ads MCP: What It Actually Lets You Do, How to Set It Up Safely, and Why I'm Not Trusting It With Live Spend Yet
Data & AnalyticsMarketing Mix Modeling Is Back: Build an MMM That Actually Works in the Post-Cookie Era
Paid Media & MarketingGoogle Demand Gen 2026: Display Ads Are Gone—Here's How to Build Full-Funnel Demand That Actually Converts
Data & AnalyticsMarketing Attribution Models in 2026: Why Last-Click Is Dead and What Replaces It
Paid Media & MarketingAgentic Commerce in 2026: When AI Shopping Bots Buy for Users — How Brands Stay Discoverable and Purchasable
Paid Media & MarketingChatGPT Ads Manager in 2026: How Conversational Intent Changes Everything About Where You Spend
AutomationAI Agents for Marketing Automation: Beyond Workflows to Autonomous Campaigns
Paid Media & MarketingReddit Ads in 2026: The Underrated Performance Channel Booming on AI and Commerce
AutomationAI Chatbots for Lead Qualification: How to Build One That Actually Filters Leads
Paid Media & MarketingGoogle AI Max for Search: The DSA Replacement and Your Migration Action Plan
Paid Media & MarketingGA4 Server-Side Tagging in 2026: How to Save Your Ad Attribution Before Cookies and ITP Kill It
AutomationMarketing Automation with n8n: How to Build a Self-Hosted Stack That Replaces $500/mo in SaaS Subscriptions
Tools & ReviewsBuild a Real-Time Marketing Dashboard: Pull 5 Ad Platforms Into One Looker Studio Report
Data & AnalyticsPredictive Lead Scoring With AI: Stop Guessing Which Leads Will Close
Data & AnalyticsAI Email Personalization Beyond Hi First Name: Real 1:1 Personalization at Scale
Paid Media & MarketingIncrementality Testing for Paid Media: Did Your Ad Actually Drive the Sale?
Tools & ReviewsAI Ad Creative at Scale: Generate Hundreds of Variations With Midjourney, DALL-E, and Runway
Paid Media & MarketingMeta Advantage+ Campaigns in 2026: When Full AI Automation Works (and When It Burns Money)
Paid Media & MarketingGoogle Performance Max: How to Actually Control the AI Black Box (Asset Groups, Audience Signals, Budget Caps)
Tools & ReviewsBuild vs Buy vs Automate: A Decision Framework for 2026
Paid Media & MarketingMeta Ads MCP: What It Actually Lets You Do, How to Set It Up Safely, and Why I'm Not Trusting It With Live Spend Yet
// 06 Questions
Honest answers.
Want me to expand on any of this? The contact page is open — no pitches.
What exactly do you do?
Two things, on purpose: full-stack engineering and performance marketing. I write code and run campaigns — most people do one. The overlap is where the leverage is.
Are you a developer or a marketer?
Both. Ten years in paid media first; I learned to code because I got tired of waiting on slow funnels. The dual skillset is the whole point.
What are you working on right now?
Always a mix: shipping this site, running paid media for a few brands, building automation pipelines in n8n, and tinkering with voice AI agents on the side. The writing page is the most current view.
Are you open to collaborations?
Selectively. If the problem is interesting and the timing fits — embedded, advisory, or paired with someone else. The contact page is the right door.
Where do I follow your work?
Here, the blog, and on GitHub, X and Substack. I write about marketing-grade engineering, automation, and the messy bits in between.
// 07 End of file
$ thanks for reading.
If anything here resonated, the contact page is open. No pitches, no funnels, no follow-up sequences.