Most of the roles I’ve held since I started working from home were with startups, where layoffs come with the territory. I used the gaps between roles to keep studying on the tech side: Salesforce certifications, business analyst training, and for the last three years, deep dives into AI. Now I’m focused on learning skills that align more directly with what I did in brick-and-mortar, but bring it full circle in an online, AI-enabled role. You can read about each stretch below.
Career History
2023 – Present Independent Studies.
Three years of self-directed study. Two parallel tracks: the Salesforce Business Analyst path and deep AI study with a portfolio of builds.
- Salesforce Certified Associate.
- Salesforce Trailhead Business Analyst path, Ranger rank (75,750 points, 82 badges).
- Creating User Stories as a Business Analyst.
- Business Process Mapping as a Business Analyst, including current state and future state stakeholder interviewing.
- Scrum and Agile methodologies.
- CRM training on Salesforce and GoHighLevel: client engagement, follow-up processes, reporting, dashboards, workflow optimization.
- Clicked: real-world Business Analyst experiences in Slack, where I became the go-to person for how to use AI to complete common BA tasks.
- Autonomee: AI-builder community (Sjoerd Tiemensma, Goda Go).
- Mark Kashef Claude Code Magic Course (19 lessons).
- Anthropic Essentials: official Claude Code training.
- Daniel Riley AI Video Boot Camp.
- Julian Goldie AI SEO.
- Course on taking an idea from concept to validated program - write the sales copy, run the ad, see what lands, then build.
- End-to-end content pipeline automation - one short description triggers a multi-stage AI workflow that generates title options, written copy, and visual assets in a single flow. Built for ad creation; the same pattern transfers to training rollouts, onboarding sequences, internal announcements, and any other multi-step content workflow that today takes hours of manual coordination.
- Claude Code automation: hooks, skills, custom workflows, and parallel-agent patterns built into a working dev environment.
- AI voice cloning and narration - trained a custom AI voice from samples and used it for product walkthroughs and page narration.
- Custom blog system built from scratch - schema-driven posts, dynamic loading, AI-search optimization.
- Motion video production for hero-level page storytelling.
- Audio-synced interactive product walkthroughs.
- SEO for AI search engines (ChatGPT, Claude, Perplexity, Google AI Overviews) - structured schema markup, llms.txt, and GEO best practices applied site-wide.
Samples scattered throughout this site.
- Ontrak Call Guide portfolio rebuild: the 2020 scripting tool, rebuilt as an AI-enhanced portfolio piece. More builds are on my projects page.
Tools Claude Code, Claude Cowork, Claude API, ChatGPT, Gemini, Convex, Supabase, n8n, Retool, Bubble, Notion, Apify, Telegram Bot API, GoHighLevel, Salesforce, Elements Cloud, Lucidchart, Figma, Canva, DaVinci Resolve, Loom…
Nov 2019 – Nov 2021 2024 – Aug 2025 Ontrak, Inc.
Ontrak phone work with health plan members on behavioral-health programs across multiple U.S. time zones. Treatment-resistant, care-avoidant populations. I carried my own book of hundreds of members and was responsible for staying in regular contact with each of them to keep them engaged in the program.
Struggling agents were assigned to shadow me. After working with several of them, I saw they were all hitting the same wall. Remote work meant there was no day-to-day pulse the way there was in a brick-and-mortar operation.
I had already been working from home as a supervisor. In the brick-and-mortar, this was exactly the work I had done as a manager and trained consultants to do. I put together a web page that helped struggling and new agents learn the call flow, all in one piece, easy to follow.
After Ontrak lost two major clients in 2021, successive layoffs cut the department from 140 reps down to 7. I was let go in the final round.
Brought back in 2024 when Ontrak was rebuilding the department in anticipation of new business.
The company went out of business permanently in August 2025.
Oct 2022 – Jul 2023 K Health.
Founding team member of the Member Activation department. Responsible for outreach calls to tens of thousands of patients across multiple U.S. time zones, qualifying them, converting them, and assisting with onboarding.
The leads came in through Google Sheets, but patient records lived in Salesforce. We had to go back and forth between the Google Sheet and Salesforce, and then back to the sheet again to make notes. It was very clunky. I taught myself Notion in a week and built the team’s call-management system inside it: today’s calls, rolling history by date with outcomes, pending follow-ups, per-person notes, lead-scoring tiers (A through D, ranked by likelihood to close), and conversion counts by tier.
The system became the department’s KPI dashboard and source of truth for conversion reporting.
Salesforce is a very popular platform, and I wanted to understand it better. I wanted to understand it well enough to have more control over the changes that would improve workflow. I also wanted to understand how a 10-minute fix could sit in the backlog as long as a 10-hour or a 10-day fix. That led me to Agile and Scrum methodologies.
The department was one of 11 closed in July 2023 when K Health’s go-to-market strategy changed significantly.
Full letter of recommendation available on request.
Sep 2016 – Jul 2019 Alliance Advisors.
The company was standing up a remote outreach department and had no internal model for how the supervisor role should look. I brought what had worked in brick-and-mortar and adapted it for a fully remote team spread across multiple U.S. time zones.
- Call-listening QA. The brunt of my day was QA. I scored agents against a quality scorecard: how fast they responded when the person picked up, how closely they followed the script, how completely they answered the questions, whether they stayed in compliance, and their tone. I calibrated so the scoring stayed consistent across the team.
During the first two weeks in this role, I discovered four team members who were hanging up on people without ever speaking to them. I could go back two weeks through their calls and never once hear their voice. It was intentional. Once I knew the pattern, I found a faster way to catch it. The first thing I checked each morning was call length, because an agent with no calls longer than a minute was almost certainly hanging up, and that is the agent I would QA first. The data showed me where to put my attention. - On-the-phones onboarding. After their initial training, new agents were assigned to me and onboarded on live calls with me for their first week.
- Group chat for team questions. Before this, questions went directly to the supervisor in DMs. Same questions, multiple agents, all day. I moved them into a group chat where every answer became a training moment visible to the whole team. In that same channel, I posted hourly stats on talk time, conversions, and more. It turned into friendly competition. Agents would DM me things like “I’m going to beat John today.”
- Task system for agent requests. Before this, agents DM’d the supervisor whenever they needed something: research the supervisor needed to run, a call note that needed correcting, a policy clarification. The supervisor had to track it, and the agent sat waiting, sometimes behind two or three other agents already waiting on the same supervisor. I set up a system inside our messaging system (Chatwork) where agents created tasks for me instead. I worked through the tasks between coaching sessions and call-listening QA. When I ticked one off, the agent got a notification that it was done. Agents stayed on the phones.
As a result of these systems, my team had the highest talk time in the company.
Mar 2014 – Jun 2017 Institute of Reading Development.
Inbound sales role selling summer reading programs to parents. My first pure phone-sales role: no face-to-face body language to read, just voice. I had to learn to listen without visual cues.
I came out of training and did not hear from anybody for about six weeks. When management finally reached out, it was to ask if I wanted to stay on for the college-classes season too. They told me what I had done straight out of training was “unprecedented.” Nobody had checked in because it was the height of the season and everybody was slammed. Their read on my performance: “if it’s not broke, don’t fix it.”
Overlapped with Alliance Advisors for my last ten months (September 2016 through June 2017). I did not return for the next IRD season because of the workload at Alliance Advisors.
Nov 2007 – Jun 2013 Jenny Craig.
Hired back to Jenny Craig in 2007 for a $50M company-wide CRM rollout. The rollout kept getting pushed back for about two years, so while I waited I ran a center in the SF Bay Area.
- Center Director, SF Bay Area. Owned all aspects of running the center, a call-heavy outreach operation that included hiring and staffing decisions, budget, inventory, training, scheduling, and the client experience. Double-digit year-over-year growth at my center while comparable locations were down as much as 30%. Outbound calls to previous clients were the engine of that growth. I reworked the call approach so consultants reached out as a genuine check-in rather than reading from the existing sales script; with the friction gone, they enjoyed making the calls and clients welcomed hearing from them. My center made 100+ previous-client calls daily while comparable locations made that volume in a full week. Jenny Craig stayed top of mind when those clients thought about weight loss again.
- Consultant management and coaching. Owned the KPIs and metrics for every consultant on my team. Set individual goals and ran weekly one-on-ones to track progress against them. When someone was off pace, I coached them on the specific areas they needed to strengthen to hit their goal. Also responsible for PIPs. Struggling consultants from other Bay Area locations were transferred to my center to be developed. It became understood that this was where someone got their strongest shot at turning their performance around, because the training and coaching I ran produced results.
- Market-level turnaround. A new role was created for me. I deployed location-to-location to streamline, declutter, and get struggling centers back on track.
- National CRM training tour. Across the U.S. and Canada. I developed the “Learn it, See it, Do it, Teach it” method for getting new centers fluent on the CRM fast. Read more about this approach →
- Back in the SF Bay Area market. Because of my fluency in the CRM, I was able to run locations simultaneously. I was also doing additional training for the West Coast market directors so they could better manage their own markets.
Laid off in June 2013 when Jenny Craig closed more than 100 locations.
Read the case study: Location Managers Were in Tears →Early Career
Before Jenny Craig: five years teaching web design at a private school in Maine, where training and building curriculum was the daily work. An earlier role at Jenny Craig in the 1990snineteen nineties as a Center Director. Management and training have been the throughline from the beginning.
Live Demos
Neither of these started as a portfolio project. Each came from a real gap a team had and couldn't close with off-the-shelf tools. Both have live demos: the Ontrak Call Guide and the K Health enrollment workflow. More builds are on my projects page.
A shared call structure for over a hundred specialists.
Ontrak's activation specialists had no shared call structure, so the ones who struggled had nothing to follow. I built one: six stages on every call, with scripts for different health conditions and coaching notes on every objection.
Open the live tool → K Health Enrollment Follow-Up.A workflow that kept a new activation team on the phone.
At K Health our enrollment leads lived in Salesforce, but it was never built for a sales team, and the change we needed was a year out in the backlog. I built a workflow in Notion that put the calendar, follow-up queue, no-show recovery, and reporting in one place. It doubled our daily outbound calls.
Open the live demo →Technology
Technology is my jam. It has always been easy for me to learn any new platform, and AI has made that easier still. If I get stuck, I screenshot it, ask AI, and work through it step by step.
- AI tools I use daily Claude (Code, Cowork, API), ChatGPT, Gemini, Grok (as needed), NotebookLM, MCP servers.
- AI image and video Nano Banana Pro, Midjourney, Higgsfield, Kling, Veo 3.1, DALL-E.
- CRM and workflow platforms Salesforce (Certified Associate, Trailhead Ranger on the Business Analyst path), Zendesk, GoHighLevel, Notion, Retool, Bubble, n8n (Railway-hosted), Make.com, Apify.
- Databases and backends Convex, Supabase (with pgvector for vector search), Postgres. I led a full Supabase-to-Convex migration across two of my projects.
- Work and collaboration Slack, Microsoft Teams, Loom, Google Workspace, Microsoft Office 365, Jira, Confluence.
- Design and content Canva, Figma, Lucidchart, Elements Cloud, Camtasia, DaVinci Resolve.
Also experienced with Cursor, Vercel, Bolt, and Windsurf, plus more than a few I have probably forgotten by now. Some of my builds are public on GitHub.