Building Gestallt: Constraints, decisions, and tradeoffs
How a documentation problem became a collaboration product—and the content strategy decisions that shaped it.
The problem
Families with gestalt language processing (GLP) children capture observations all the time— but in scattered ways. Texts to the SLP. Notes in the phone. Scraps of paper that get lost. The SLP sees maybe 1–2 hours per week. The other 166 hours produce language that never makes it into therapy planning.
The existing solutions didn't fit:
- Paper logs: Not searchable, easy to lose, hard to share
- Shared docs: No structure, no stage tracking, HIPAA concerns
- EMR systems: Parent can't contribute, clinician-only access
- AAC apps: Built for communication, not documentation
- Generic note apps: No GLP framework, no team permissions
The gap: a tool that lets parents capture quickly, gives clinicians structure, and keeps everyone on the same page with appropriate privacy.
Constraints I worked with
Two audiences, different needs
Parents need speed and simplicity. Clinicians need structure and clinical fields. The same entry needs to serve both—without making either audience uncomfortable.
Privacy without friction
Sensitive data requires team isolation and role permissions. But the signup/invite flow can't be so complex that families give up before capturing their first entry.
HIPAA language accuracy
Clinicians will ask about HIPAA. I can't overpromise "compliance" (requires organizational policy) but need to credibly describe the infrastructure's capabilities.
Non-deficit framing
Autism documentation often uses clinical/deficit language. The product and content needed to feel supportive—celebrating progress, not pathologizing behavior.
Key decisions
Two modes, one record
Rather than separate "parent view" and "clinician view," both modes write to the same entry. Parents capture with Flexible Mode (phrase, context, tags). Clinicians add structure with Clinical Mode (stage, function, notes). Neither mode blocks the other.
Why: Reduces data fragmentation. The SLP sees parent observations in context. Parents see clinical notes without needing a separate system.
Child-centered teams, not user-centered accounts
The team is organized around the child, not the parent. This means professionals supporting multiple families can switch between teams with one account, and data isolation happens at the team level.
Why: Matches how care actually works. Dr. Chen doesn't need 12 accounts for 12 families. Privacy boundaries are at the team, which is where they matter.
"HIPAA-compatible infrastructure" language
I deliberately avoided claiming "HIPAA compliant" or "HIPAA certified" (no such thing). Instead, documentation describes the technical controls (team isolation, verified membership, Firebase/Google Cloud infrastructure) and what they support.
Why: Honesty builds trust. Clinicians who know HIPAA will see through overclaims. Accurate language about infrastructure vs. organizational policy is more credible.
GLP stages as optional structure
Stage tracking (1–6) is available but not required. Parents unfamiliar with GLP stages can ignore them. SLPs who want them can add them. The data model supports both.
Why: Reduces barrier to entry for families. Doesn't force clinical knowledge on parents. Keeps the tool usable even if the family isn't doing formal GLP work.
Tradeoffs I accepted
Not an AAC replacement
Gestallt is documentation, not communication. I could have built AAC features, but that's a different product with different expertise. Better to do one thing well.
No offline mode (yet)
Real-time sync is simpler to build and maintain. Offline-first would require conflict resolution logic. I chose connectivity as a requirement for v1.
No built-in analytics
Progress visualization would be valuable, but filtering by stage/date gives clinicians what they need. Charts can come later; core capture/share comes first.
English-focused for now
Localization would expand reach, but GLP resources are primarily English. Better to serve the core audience well before expanding.
What shipped
The core product: a collaborative phrasebook with two documentation modes, team-based permissions, GLP stage tracking, and HIPAA-compatible infrastructure.
The content system: positioning, messaging, objection handling, competitive analysis, and ready-to-send sequences for recruiting, clinician outreach, and community engagement.
The documentation: accurate, careful language about security posture. Non-deficit framing throughout. Pages that answer buyer questions in the right order.
Content strategy artifacts
This microsite demonstrates the content system I'd build for any product launch. Here's what's included:
Messaging house →
3 pillars, benefit statements, proof points, and sample headlines by audience.
Objection handling →
Talk tracks for "we already use notes," privacy concerns, time burden, and clinical adoption.
Competitive positioning →
How Gestallt compares to paper logs, shared docs, EMR notes, AAC apps, and generic note apps.
Documentation →
Concepts, data model, security posture, architecture, limitations—docs that build trust.
This is what I do
Product-focused content strategy: positioning, messaging, docs that convert, objection handling, and content systems that scale with product velocity.