Hi, I'm Alec Rothman
I'm a data leader with 15+ years turning complex problems into measurable outcomes, leading analytics and product teams during the week, and shipping projects on weekends. Equally at home writing SQL, debugging something, and presenting to a CMO.
About me
I'm a Data & Analytics leader with 15+ years turning data into measurable outcomes. Currently leading Salesforce's marketing data infrastructure, I combine technical expertise with creative problem-solving - from shipping tools that accelerate marketing execution to building dashboards that actually get used.
My core strength is understanding and connecting data, from raw collection to curation and taking action. I excel at bridging the gap between complex data systems and business needs. Whether it's harmonizing identity resolution across disparate sources and grains or designing GDPR-compliant analytics across global properties, I love solving the tricky problems that make data actually useful.
On weekends, I ship solo projects end-to-end using AI and LLMs. Projects like SheetTalk and Blunt Therapist showcase my belief that AI should solve real problems while maintaining user trust. Staying hands-on makes me a better leader.
Technical Toolkit
My professional experience
Salesforce
2018 - PresentDirector of Data & AI Products
May 2023 - PresentCurrently leading teams building Salesforce's Marketing Data Warehouse, delivering dozens of foundational Snowflake datasets with dbt, Airflow, and semantic layers. That includes canonical data models connecting billing, usage, and pipeline data to marketing engagement across the funnel, and the data tools built on top that have become sources of truth for product and field marketing strategy. We've shipped AI analytics pilots and brought rapidly developed prototypes to production that accelerated marketing execution by 50%+!
I spend my days juggling business needs with practical execution, making sure we're delivering data products with ~100% uptime and real trustworthiness. I stay hands-on and accelerate contributions through evolving AI-powered development and delivery cycles.
Director of Analytics
February 2022 - April 2023Built data science feature stores that revealed 77% of $Billions of Annual Order Value growth came from customers engaging with learning and community platforms - and those customers showed 380% higher median AOV. Also unified 20M+ user profiles across learning, community, and developer platforms to help us understand how people were actually using our platform. These insights got shared on public company earnings calls! Managed a team of analysts and data engineers who maintained all sorts of data goodies: pipelines in Airflow, SQL data models, and Tableau dashboards.
Director of Marketing Technology (Heroku)
August 2020 - January 2022Built a custom cross-domain web analytics platform across a dozen Salesforce properties (4.5M+ monthly visitors), revealing massive digital traffic overlap for US-based companies. Integrated predictive modeling with marketing platforms to automate high-value customer nurture programs at the moment of signup. Got my hands dirty designing cross-domain experiences that achieved a 151% increase in signups for targeted developer programs, and led experiments that delivered a 7% lift in platform payment account creation based on product signals.
Marketing Technology Manager (Heroku)
December 2018 - July 2020Started my Salesforce journey at Heroku, where I owned the marketing tech strategy. Developed a predictive algorithm that outperformed third-party services by 89% for detecting potential abuse at signup - Security loved it! Built Looker dashboards and LookML semantic models to visualize acquisition performance across 100K+ monthly signups.
Autodesk
2014 - 2018Principal Analytics Engineer
2016 - 2018Architected a scalable, GDPR-compliant enterprise web analytics platform serving 20M+ monthly visitors and 90M+ monthly page views across global websites. Here's the cool part: I built the GDPR consent logic using Tealium that worked seamlessly across all of Autodesk's top-level domains, making sure user privacy preferences were respected globally, and it's still in use today! Also got to play in the Account Based Marketing space, setting up real-time use of first-party data for personalization.
Senior Technical Web Analyst
2014 - 2016Led the charge on foundational web analytics platform and tag management migrations and upgrades, making it more scalable and future-proof. Also played tech matchmaker, integrating services like Google Analytics and Adobe Analytics as well as data enrichment, optimization, and voice-of-customer platforms into our analytics ecosystem.
CREDO Mobile
2011 - 2013Web / Marketing Data Analyst
Built a regression model for predicting customer churn that outperformed the existing model by 10% in a friendly 30-day competition. Implemented a remarketing display ad campaign with A/B testing that drove a 15% conversion lift. Also expanded the customer matching framework using advanced SQL and Levenshtein distance to make sure we weren't accidentally showing acquisition offers to existing customers.
Greenpeace USA
2008 - 2011Data Analyst (organization's first)
Pioneered the organization's first dedicated data analysis role, managing insights from ~1.5M supporter records across fundraising, advocacy, email communications, and grassroots initiatives. Developed KPI reporting for a nationwide canvassing program that generated 1M+ face-to-face contacts and recruited 75K+ monthly donors annually. This is where my journey in data analytics began, working for a cause I believed in while learning the ropes of turning numbers into insights.
Some personal projects
SheetTalk
Chat with your spreadsheets using natural language while keeping your data completely private - nothing leaves your device. Upload a CSV or Excel file and start asking questions in plain English.
Built with Claude Code to explore how AI can provide powerful insights while respecting user privacy.
- Privacy-first architecture - your data stays local, no server-side storage
- Natural language queries - ask questions about your data in plain English
- AI that understands context and suggests follow-up analyses
Blunt Therapist
An AI grounded in real therapeutic principles but delivers advice with refreshing directness. It's like having a brutally honest friend who also happens to be a trained professional.
Built with a PhD psychologist to explore how LLMs can be used responsibly in sensitive domains.
- Built with Anthropic's Claude (pre-Claude Code) for authentic therapeutic interactions
- Privacy-first - no conversation data stored
- Safety mechanisms designed for sensitive mental health contexts
Weathervane
Behavioral analytics without a vendor. Weathervane observes user behavior and emits structured events directly in the browser - no backend, no dashboards, no lock-in.
Built to solve the problem of duplicated tracking logic and vendor SDK lock-in. Applications emit behavior once; destinations become a deployment detail.
- Tracks content exposure, form engagement, rage clicks, errors, and web vitals
- ~8KB gzipped, zero dependencies, zero cookies, zero network requests
- Forward events to GA4, PostHog, Mixpanel, or your own API - all at once
Menu Scanz
This was an exploration of AI, and also trying to solve a personal problem: dietary restrictions at restaurants. I wanted to snap a photo of a menu and understand if any items would be problematic for me.
A weekend project where I learned firsthand about integrating vision APIs with LLMs.
- Vision AI extracts and understands menu items from photos
- Personalized recommendations based on your dietary needs
- Instant analysis - no manual menu reading required
Long Pong
The asynchronous ping pong game for procrastinators and busy people. Take up to 7 days per turn because good things come to those who wait! A turn-based online game that removes the pressure of real-time gameplay.
Built almost entirely with Claude Code to continue exploring product planning and development with AI assistance, leveraging detailed PRDs and build plans.
- Developed comprehensive product requirements and technical specifications with AI
- Built full-stack multiplayer game using Claude Code for nearly all development
- Explored AI-assisted product planning and development workflows
Cheeky Stuff
A "cheeky" NFT project combining technical innovation with humor. I developed a custom PNG-to-GIF composition system that could generate seamless animations from hand-drawn frames and combine traits with configurable rarity settings. This powered the creation and mint-out of 10K unique NFTs, each with their own personality and characteristics.
Developed all the code myself and managed the all the launch aspects and minting dynamics at scale.
- Built custom GIF composition engine that could seamlessly combine 6-8+ hand-drawn PNG frames with configurable rarity tiers
- Developed Ethereum smart contract and Node.js web app with comprehensive analytics tracking (14.4K+ events captured)
- Handled viral traffic spikes with 3.4K users across 11.5K sessions
- Successfully tracked full minting funnel: 5.2K wallet connects led to 2K+ completed mints