01
The win: first synthesis
Run a starter synthesis prompt on one pasted thread of your real data in the first half hour. Themes, severity, sources — visible immediately.
Done looks like A themed output from your own thread.
Workshop · 2 live sessions
Your scattered signals, turned into cited PRDs — on your own data.
Customer signal lives everywhere — Slack, call recordings, support tickets, meeting notes — and one person always ends up as the bottleneck synthesizing it. Build the pipeline that fixes this, on your real sources, and ship one output to a real stakeholder by the end.
Taught 2,000+ professionals — cohorts, fellowships, teams
Start here
What's a product context pipeline — and what changes when you have one?
Never heard the term?
Customer signal lives everywhere: Slack threads, call recordings, support tickets, meeting notes. A context pipeline is the setup that gathers it safely and turns it into synthesis with citations — themes, severity, root causes, every claim clickable back to its source.
GIF placeholder · Slack export → notes repo → cited synthesis
Already the person who synthesizes?
Then you are the pipeline, and you know the cost: every analysis starts from zero and your week disappears into reading. With a pipeline, the machine clusters and cites while you validate and sign — and the notes repo means the next analysis starts smarter than the last.
The synthesis is automated. The judgment isn't.
Either way you leave the same: one output shipped to a real stakeholder, every claim cited, from a pipeline that keeps running weekly after the workshop.
You leave with
Three artifacts, all yours, all built on your data.
- A working pipeline on your own data: exports, notes repo, synthesis skill
- One output shipped to a real stakeholder, every claim cited to source
- The prompt/skill kit to keep running it weekly
Who it's for
For the person who ends up synthesizing everyone's signal by hand.
A good fit
- PMs, founders, researchers, and design leads
- Teams where one person is the bottleneck turning Slack, calls, and tickets into decisions
- Non-technical is fine. Cowork is the gentler path; Claude Code goes further on automation
Not for you if
- You have no access to customer signal — no Slack, support queue, calls, or notes to draw from
- You want fully automated synthesis with no human in the loop — the judgment is the point
- You need real-time analytics infrastructure, not a weekly synthesis pipeline
Bring
- Access to your own Slack, support tickets, or notes — your workspace, your permissions
- One stakeholder who'd receive the output
- The free Claude Cowork intro course, done
- The published write-up, read — it's the syllabus
Agenda
Session one: sources and context. Session two: synthesis and shipping.
Session one · 2.5 hours
Show & tell: sources & context
Watch every block run live on real data, end to end.
Between sessions · one week
Your turn
Your pipeline runs on your own sources for a week — homework, with help in the community.
Session two · 2.5 hours
Bring your build: synthesis & shipping
Session two starts from what your pipeline gathered.
Every session is recorded
Session one · Watch it run: sources and context · 2.5 hours
02
Map your signals, export safely
Score your sources by value and accessibility, pick the top two, and export the first as structured data — session-based, nothing leaves your org.
Done looks like A signal map and your first export.
03
The common notes repo
Stand up the cumulative context layer: customer background, known pains, terminology. Seed it with what you already know — it compounds from here.
Done looks like A notes repo that makes the next analysis smarter.
Between sessions · Your turn, one week
Your pipeline accumulates a week of real use — that week is what proves the context compounds instead of resetting. Run your exports, feed the repo, note where it fights you, and bring the questions; stuck points get answered in the community along the way.
Session two · Bring your build: synthesis and shipping · 2.5 hours
04
Write your synthesis skill
Structured prompts that demand themes, severity, frequency, citations, root causes, and proposed fixes. Run it over both exports plus the repo.
Done looks like Your synthesis skill, running on your data.
05
Judge the output
Review one synthesis run the way a human should: validate, correct, annotate. What the machine clustered is a draft; what you sign is the deliverable.
Done looks like A validated synthesis with your judgment visible.
06
Format and ship
Shape the output for its reader — PRD input, knowledge-base article, or an agentic customer-success doc — and send it to the real stakeholder before the session ends.
Done looks like Shipped, with clickable citations.
Proof
Every method has a receipt.
- The context pipeline, written up — the full workflow, built and run in production, published free. It's also the syllabus
- github.com/iruhdam1/skills — the skills, public
- What participants say — the shared wall of love, filled from cohort 1 onward
Testimonial · name · role
Testimonial · name · role
Testimonial · name · role
Instructors
Taught by people who ship this way daily.
Photo
Madhuri Maram
Lead instructor. Product designer, 2,000+ professionals trained across cohorts and fellowships, designing agentic products in production today.
Photo
Arjun
Co-instructor. Building the course material with Madhuri. Bio and receipts land on the instructors page.
Price & rooms
One price. Pick your room, then hold your seat.
$499 per seat · October cohort · live online, recorded. Both rooms keep the week between sessions. First room to reach eight runs; dates set at least two weeks ahead. Your pick pre-fills the WhatsApp message.
- Two live sessions of 2.5 hours, one week apart, on your own data
- Homework and help over the community between sessions
- Recordings of every session
- The prompt/skill kit, yours to keep running weekly — plus the free Claude Cowork intro course and write-up as pre-work
FAQ
Questions, answered.
Is my company's data safe in this?
You export through your own workspace permissions and everything stays on your machines — the method uses no external servers. A consent-and-redaction checklist is part of the pre-work, and what you shouldn't export is covered explicitly.
What do I need before session one?
Access to your own Slack/support/notes, one stakeholder who'd receive the output, the free Claude Cowork intro course, and the published write-up — it's the syllabus.
Why is this two sessions a week apart?
The notes repo needs a week of real use between sessions — that's what proves the context compounds instead of resetting.
Claude Code or Cowork?
Either works; the session shows both. Cowork is the gentler path for non-technical folks, Claude Code goes further on automation.
Why $499?
It's the introductory-cohort price. The first cohorts shape the format, and the price says so. It rises for later cohorts.
What if the live time doesn't fit my timezone?
Every session is recorded, and the work happens between sessions on your own time. Follow the recording and bring questions to the community.
Seats
October rooms are forming.
Pick your room above — the WhatsApp message pre-fills.