Customer Discovery Synthesizer
Attach interview notes and a research agent synthesizes evidence-backed patterns, attributed quotes, and a Jobs-to-be-Done map into a living discovery doc, proposes follow-ups, and verifies every finding against its source before you approve.
Saves you ~4.4 h / run
How it works
- Trigger
- When you start the “Synthesize discovery” workflow.
- Job
- Read notes, extract signal, map JTBD, update the doc, verify against source, and approve.
- Outcome
- A sourced customer-insight brief and updated JTBD map.
What it installs
Agents 2
-
Discovery Synthesizer
Turns raw interview notes into ranked patterns, attributed quotes, and a JTBD map.
-
Customer Discovery Synthesizer Quality Reviewer
Checks primary evidence, domain controls, deliverable completeness, and communication quality, stopping the run when the work is wrong, unsupported, incomplete, or uncertain.
Teams 1
-
Customer Discovery Synthesizer quality team
The delivery agents produce the work while an independent quality reviewer checks each workflow handoff against explicit evidence, domain, and communication requirements before the run can continue.
Workflows 1
-
Synthesize discovery
Read notes, extract signal, map JTBD, update the doc, verify against source, and approve.
Documents 1
-
Discovery doc
The living synthesis the agent maintains and updates with each batch of notes.
Goals 1
-
A living, evidence-backed model of the customer that stays current
Discovery exists so customer learning compounds instead of dying in scattered notes — every batch of interviews sharpens one durable, trustworthy model of who the customer is and what job they're hiring the product for. Success looks like: The discovery doc reflects every batch of attached interviews, where each pattern is backed by ≥2 interviews and an attributed quote, the JTBD map separates jobs from solutions, surprises and assumption-breaks are surfaced, confidence is labelled honestly against the sample size, and no finding is unsupported by the source notes.
Skills 3
-
conducting-user-interviews
Read discovery conversations like a researcher: collect stories not opinions, weight switching over bitching, falsify rather than validate, right-size conclusions to the sample, and surface surprises. Adapted from refoundai/lenny-skills/conducting-user-interviews.
-
summarize-interview
Transform an interview transcript into a structured summary focused on JTBD, current solution, likes, problems, key insights, and dated action items — in plain language, using verbatim quotes. Adapted from phuryn/pm-skills/summarize-interview.
-
jobs-to-be-done
Apply the structured JTBD lens — functional/social/emotional jobs, pains (challenges, costliness, mistakes, unresolved), and gains (expectations, savings, adoption, life improvement) — separating the job from the solution and ranking by intensity. Adapted from deanpeters/product-manager-skills/jobs-to-be-done.
Folders 1
-
Customer Discovery
Requirements
- Interview notes or transcripts attached — The agent synthesizes from the interview notes, transcripts, or recordings you attach. The more verbatim the source — actual quotes and stories rather than your paraphrase — the stronger the synthesis, because every finding must be traceable to a real line a customer said.
Setup guide
How to Synthesize Customer Discovery Interviews
A practical guide to turning interview notes into evidence-backed patterns, quotes, JTBD maps, and follow-up tasks.
Read the setup guideDon't see your workflow? Describe it.
A sentence or two about a recurring job is enough. We design the playbook that runs it and show you exactly what it saves.