Synapse · Design Partner & SME Program

The Vision

Intelligence that becomes your co-founder

The Intelligence module discovers, structures, enriches, and refines everything about your venture. A living system that compounds intelligence with every interaction.

Feedback → Product Pipeline

How Feedback Becomes Shipped Improvements

Design Partners

Our Founding Partners

10Demo

Validation

Articos

Ideation

PureVPN

Growth

Paio

Traction

Zamanat

Validation

Who's Involved

Two Distinct Roles, One Mission

Design Partners and Subject Matter Experts work in tandem — but never in the same session. This ensures unbiased feedback and expert-grade improvement documents.

🎯

Design Partners

Real founders using Synapse with their own ventures. They represent the authentic user voice — testing features in context, surfacing friction through genuine usage.

Real founders at different venture stages (Ideation → Growth)
Participate in structured test sessions: usability tests, guided reviews, workshops, interviews
Provide feedback on their own ventures' data — not synthetic examples
NPS scored at every session close to track sentiment over time
Never told what to think — observed in how they navigate and react

🧠

Subject Matter Experts

Domain specialists assigned per feature. Never present during DP sessions. After each session, compiled feedback is handed to the SME who writes the Improvement Requirement Document (IRD).

Domain experts in AI quality, UX, market intelligence, analytics, and more
Receive raw recordings, transcripts, and timestamped notes — not summaries
Write IRDs that translate friction into specific, actionable improvement hypotheses
Each SME owns specific features and is accountable for their improvement arc
Coordinate retests after fixes ship to confirm hypotheses

Tooling

Testiment — Testing & Feedback Platform

Testiment powers the entire DP & SME feedback loop with two core functions: Bug Reporting and Test Management.

🐛Bug Reporting

Testers report bugs with a single click — the system records the screen session and automatically collects all context engineers need.

Screen recording captured on submission
Console log captured automatically
Step-by-step reproduction path
Network tab activity snapshot
Browser & device metadata
Direct link to the exact state

🧪Test Management

Structure and run tests for Design Partners — NPS, think-aloud recording, and page-by-page guided tasks compiled for SME review.

NPS surveys with automatic scoring & tracking
Think Aloud — record and talk through your experience
Page-by-page tasks the tester completes step by step
All session data collected and structured for SME handoff
Results tagged with ref codes for full traceability

End-to-End Pipeline

From Bug Report to Shipped Feature

Bug & Test

DPs report bugs & run tests

→

Sessions Recorded

Screen, console, steps, network

→

AI Ingests

Results auto-structured

→

SME Review

Solution via IRD

→

IRD Submitted

Issue + hypothesis + priority

→

PO Review

Accept or refine

→

Agents Build

AI agents implement fix

Sample IRDs

What Improvement Requirement Documents (IRDs) Look Like in Testiment

Each IRD captures the issue, the SME's hypothesis for fixing it, what they did to validate, and the results. Use the arrows to browse examples.

1 / 4

IRD-001 · Canvas summary lacks competitive context

P0 — Critical

Ref Code

P1-IN-F1-A2-10DEMO-UT-SME01

Feature

F1 — Canvas: Venture Overview & Research Summary

Feedback Area

A2 — Output Quality

DP / Method

10Demo (Validation) · Usability Test

SME

SME-01 — Output Quality & Methodology

Issue

Canvas summary does not reference competitor positioning or market signals, making the output feel generic and low-confidence to the founder.

Hypothesis

Augment the canvas agent prompt with competitor signal injection at generation time. With access to top 3 competitor profiles and latest market signals, the output will include competitive context founders find actionable.

Validation Method

What SME Did

Manually injected top 3 competitor profiles + market signals into the canvas agent prompt for 10Demo's venture. Ran 3 generations with and 3 without injection, then had the DP blind-rate outputs on a 1–5 actionability scale.

Test Scope

A/B comparison across 6 canvas outputs (3 control, 3 augmented) with the same venture context.

Hypothesis Results

✓ Validated

Actionability score

2.1 → 4.3 +105%

✓ Validated

Competitive mentions

0 → 3.7 avg per output

● Noted

Generation latency

+1.8s (acceptable)

Conclusion

Hypothesis confirmed. DP rated augmented outputs significantly more actionable. Competitor injection added ~2s latency, within acceptable range. Approved for Sprint 1.

IRD-002 · Market research monitor shows stale data

P1 — High

Ref Code

P1-IN-F2-A8-ARTICOS-AS-SME02

Feature

F2 — Market Research Monitor

Feedback Area

A8 — Reliability & Trust

DP / Method

Articos (Ideation) · Async Survey

SME

SME-02 — Market Intelligence & Data Accuracy

Issue

Market research dashboard displays data from the latest run but doesn't indicate when it was generated. Founders can't tell if insights are 2 days or 2 months old, eroding trust in the output.

Hypothesis

Add a "Last researched X days ago" badge + freshness indicator (green/amber/red) to every data card. If data is >14 days old, surface a prompt suggesting a refresh. This will increase trust scores.

Validation Method

What SME Did

Created a Figma prototype with freshness badges on all data cards. Ran a 5-second test with Articos and PureVPN — showed them the current UI vs the prototype side by side. Asked: "Which one do you trust more?"

Test Scope

2 DPs (Articos, PureVPN). Side-by-side comparison. Follow-up NPS on data trust.

Hypothesis Results

✓ Validated

Trust preference

2/2 DPs chose prototype

✓ Validated

NPS trust sub-score

5 → 8 (out of 10)

✓ Validated

Refresh click intent

Both said they would refresh

Conclusion

Freshness badges are a clear trust signal. Both DPs immediately preferred the badged version. Refresh prompt drives engagement. Approved for Sprint 1.

IRD-003 · ECP scoring feels opaque to founders

P1 — High

Ref Code

P1-IN-F6-A4-PUREVPN-INT-SME04

Feature

F6 — Target Audience: ECP / ICP

Feedback Area

A4 — Agent Interaction

DP / Method

PureVPN (Growth) · In-depth Interview

SME

SME-04 — AI / Agent Quality

Issue

ECP composite scores (0–100) appear as a single number with no breakdown. The founder at PureVPN said "I don't know why this segment scored 72 vs 58 — I can't challenge it or trust it."

Hypothesis

Expose the 6 criteria sub-scores (pain severity, WTP, accessibility, urgency, market size, frequency) as a spider/radar chart next to each ECP card. Add a "Why this score?" expandable section with evidence citations.

Validation Method

What SME Did

Built a clickable prototype with spider charts and expandable evidence panels. Ran a cognitive walkthrough with PureVPN and Secure.com — asked them to compare two ECPs and explain which one they would prioritize and why.

Test Scope

2 DPs (PureVPN, Secure.com). Cognitive walkthrough + think-aloud. 3 ECP cards each.

Hypothesis Results

✓ Validated

Ability to compare ECPs

"Night and day" — PureVPN

✓ Validated

Trust in scoring

3/10 → 8/10 avg

▲ Partial

Evidence citations

Useful but "too academic"

Conclusion

Spider chart is a clear win. Evidence panel needs simplified language — less academic, more founder-friendly. Approved for Sprint 2 with copy revision task.

IRD-004 · Doctrine onboarding has no guided path

P2 — Medium

Ref Code

P1-IN-F7-A9-WELLOWS-WS-SME01

Feature

F7 — Doctrine: Venture Docs

Feedback Area

A9 — Onboarding & Learnability

DP / Method

Wellows (Ideation) · Workshop / Co-design

SME

SME-01 — Output Quality & Methodology

Issue

First-time users land on an empty 16-folder vault with no indication of where to start. Wellows spent 4 minutes clicking between folders before asking "what do I do first?" Dropout risk is high at this moment.

Hypothesis

Add a "Getting Started" guided tour that highlights 3 recommended first actions: (1) populate the Concept folder, (2) run your first market research, (3) chat with Neo about your venture. Show a progress tracker in the sidebar.

Validation Method

What SME Did

Designed a 3-step guided tour overlay in Figma. Co-designed the flow with Wellows and Articos in a workshop session — iterated on step copy, placement, and whether it should auto-dismiss or require manual close.

Test Scope

2 DPs (Wellows, Articos). Workshop co-design + prototype test. 2 rounds of iteration.

Hypothesis Results

✓ Validated

Time to first action

4 min → 45 sec

✓ Validated

Completion rate

100% finished 3 steps

▲ Partial

Auto-dismiss preference

1 wanted auto, 1 wanted manual

Conclusion

Guided tour dramatically reduces time-to-first-action. Default to manual dismiss with a "Don't show again" option. Approved for Sprint 2.

Master Rollout

Phase-by-Phase Feature Testing

Every feature tested with specific methods, assigned to SMEs, tracked through a complete feedback lifecycle.

P1Intelligence7 features · 23 Mar

▾

Code	Feature	Feedback Areas	Methods	SME
F1	Canvas — Venture Overview	UX · Output Quality · Data Presentation · Feature Value · Time Saved · Methodology	UTGRNPS	SME-01
F2	Market Research — Monitor	UX · Output Quality · Data Presentation · Feature Value · Time Saved · Reliability	UTASLOG	SME-02
F3	Competitor Research — Monitor	UX · Output Quality · Data Presentation · Feature Value · Reliability · Missing Cap.	UTINTAS	SME-02
F4	Product Research	UX · Output Quality · Data Pres. · Feature Value · Time Saved · Methodology · Reliability	UTGRNPS	SME-02
F5	Brand Page	UX · Data Presentation · Feature Value · Reliability · Onboarding	5SUTNPS	SME-03
F6	Target Audience — ECP / ICP	UX · Output Quality · Agent Interaction · Feature Value · Methodology · Reliability	UTINTCOG	SME-04
F7	Doctrine — Venture Docs	UX · Agent Interaction · Feature Value · Methodology · Onboarding · Stage Approp.	UTWSLOG	SME-01

P2Analytics9 features · 1 Apr

▾

Code	Feature	Methods	SME
F1	War View	UTGRLOG	SME-06
F2	Theatre View	UTASLOG	SME-06
F3	Battles View	UTINTNPS	SME-07
F4	Battle Zones	UTCOGLOG	SME-07
F5	Journeys View	UTWSAS	SME-06
F6	Profiles View	UT5SNPS	SME-03
F7	Strikes	UTASLOG	SME-07
F8	Integrations	UTASLOG	SME-11
F9	Event Mapping	UTCOGGR	SME-11

P3Actions5 features · 8 Apr

▾

Code	Feature	Methods	SME
F1	Strike Validation	UTINTNPS	SME-07
F2	Atomic Actions	UTCOGLOG	SME-04
F3	Squad Setup	UTGRNPS	SME-04
F4	Troops	UTINTLOG	SME-04
F5	PowerUps	UTASNPS	SME-11

P4Konnect5 features · 15 Apr

▾

Code	Feature	Methods	SME
F1	Contacts	UTASLOG	SME-09
F2	Outreach	UTA/BLOG	SME-09
F3	Inbox	UT5SNPS	SME-09
F4	Calendar	UTINTNPS	SME-09
F5	Interview Toolkit	UTWSINT	SME-09

P5Build2 features · 22 Apr

▾

Code	Feature	Methods	SME
F1	Spec Generation	UTCOGINT	SME-10
F2	Product Builder	UTINTLOGNPS	SME-10

SME Registry

Domain Expert Assignments

Each SME owns specific features and writes IRDs that translate friction into actionable hypotheses.

SME-01

Output Quality & Methodology

P1-IN-F1, P1-IN-F7

Prompt tuning, agent output quality, stage-gating

SME-02

Market Intelligence & Data Accuracy

P1-IN-F2, F3, F4

Research output accuracy, data completeness

SME-03

UX & Onboarding

P1-IN-F5, P2-AN-F6

Onboarding flows, learnability, first-impression clarity

SME-04

AI / Agent Quality

P1-IN-F6, P3-AC-F2, F3, F4

HITL, agent trust, directive quality

SME-05

Knowledge Mgmt & Stage-Gating

P1-IN-F7

Folder structure, agent-to-folder mappings

SME-06

Analytics & Data Visualisation

P2-AN-F1, F2, F5

Charts, data density, cross-theatre consistency

SME-07

Strike Intelligence & Battle Zones

P2-AN-F3, F4, F7, P3-AC-F1

AAARRR mapping, lifecycle transitions

SME-08

Outreach & Messaging Quality

P3-AC-F5 (Outreach PowerUps)

Copy quality, A/B signal interpretation

SME-09

Konnect & Outreach Intelligence

P4-KO-F1 through F5

Contact enrichment, scheduling, synthesis

SME-10

Build & Prototype Quality

P5-BD-F1, F2

SDD accuracy, scaffold success rate

SME-11

Integrations & Data Connectivity

P2-AN-F8, F9, P3-AC-F5

Connector coverage, event mapping accuracy

Taxonomy

Test Methods, Feedback Areas & Reference System

The complete classification system for collecting, categorizing, and tracing every piece of feedback.

10 Test Methods

Usability Test

Live moderated task walk-through

Guided Review

Demo + structured Q&A

Async Survey

Structured feedback form

INT

In-depth Interview

1:1 deep-dive session

5-Second Test

First-impression snap test

COG

Cognitive Walkthrough

Expert heuristic evaluation

LOG

Usage Log Analysis

Behavioural data from sessions

A/B

A/B Experiment

Variant comparison test

NPS

NPS / CSAT Pulse

Post-session satisfaction

Workshop / Co-design

Collaborative design session

12 Feedback Areas

User Experience

Navigation, cognitive load, hierarchy

Output Quality

Accuracy, relevance, completeness

Data Presentation

Charts, tables, info density

Agent Interaction

Directives, HITL friction, trust

Feature Value

ROI, JTBD fit, willingness to pay

Time Saved

Speed vs manual, efficiency delta

Methodology

Prompt structure, step sequencing

Reliability & Trust

Consistency, confidence in outputs

Onboarding

Time-to-value, discoverability

A10

Integration Fit

Workflow compatibility, data I/O

A11

Stage Appropriateness

Right feature at right stage

A12

Missing Capability

Gaps vs JTBD not served

Reference Code Anatomy

P1-IN-F3-A4-10DEMO-UT-SME2

Phase 1

Intelligence

Competitor Research

Agent Interaction

10DEMO

Design Partner

Usability Test

SME2

Assigned SME

Building IntelligenceTogether

10Demo

Articos

PureVPN

Paio

Zamanat

Design Partners

Subject Matter Experts

🐛Bug Reporting

🧪Test Management

IRD-001 · Canvas summary lacks competitive context

IRD-002 · Market research monitor shows stale data

IRD-003 · ECP scoring feels opaque to founders

IRD-004 · Doctrine onboarding has no guided path

10 Test Methods

12 Feedback Areas

Reference Code Anatomy

Building Intelligence
Together