What Makes Us Different
How we approach transparency, accuracy, and ethical design in ancestor personas
This page explains the principles, technical foundations, and ethical commitments behind how ThroughTheirEyes.ai approaches ancestor personas with transparency and care.
Many general-purpose AI tools are tuned to keep the conversation flowing, even if that means guessing or filling in gaps without telling you where those details came from. That can be fine when you’re just looking for a good story—but it’s risky when you’re dealing with family history. In ThroughTheirEyes.ai, we still care deeply about vivid, in-world realism, but we make the boundaries visible: you can see which parts are grounded in your records and which are careful inferences drawn from historical context.
ThroughTheirEyes.ai was built differently from the ground up.
Our Core Principles
1. Provenance-First Architecture
Every factual statement in our platform includes a provenance marker—a citation showing exactly where that information came from.
Example conversation:
"My father Griffith was a miner, like his father before him." [P1]
Source: [P1]
GEDCOM occupation field
Confidence: 95%
This transparency ensures users can validate every claim and understand the difference between verified genealogical facts and contextual inference.
What Is Provenance Tracking?
Provenance tracking means every factual claim contains a reference to its source—GEDCOM records, user input, or contextual inference. This ensures complete transparency and prevents AI fabrication.
2. Synthetic-Only Demos
William Price, our demo persona, is a fully synthetic ancestor—not based on any real individual.
This protects privacy and ensures no real family's genealogy is exposed during demonstrations. Every detail reflects authentic 19th-century Welsh mining communities, but zero names, dates, or relationships correspond to actual people.
Why this matters:
Other AI tools often demonstrate features using real historical figures or crowdsourced family trees. We believe that approach compromises privacy and sets the wrong ethical precedent.
What Is a Synthetic Ancestor?
A synthetic ancestor is a fictional individual whose historical details, family structure, and cultural context are plausible and historically accurate, but not derived from real people. This approach protects privacy while demonstrating how AI genealogy works.
3. Blueprint-Based Persona Construction
Each ancestor persona is built from a structured blueprint containing only verified genealogical data. This includes:
- ✓Names, dates, places from GEDCOM records
- ✓Family relationships and occupations
- ✓Historical context enrichment (optional)
- ✓Hard cutoff dates (ancestors can't know events after their death)
The AI operates within these strict boundaries—it can use cultural context ("miners often sang hymns") but never invents personal events. Conversations rely on verified genealogical data plus short-term conversational context (not persistent memory).
What Is a Persona Blueprint?
A persona blueprint is a structured profile containing only verified genealogical data and historical context. It defines strict boundaries for what an AI ancestor can and cannot know, preventing fabrication while enabling natural conversation.
4. Your Data Stays Yours
We never train AI models on your family tree. Period.
Encrypted at rest
Your GEDCOM files are encrypted using industry-standard AES-256 encryption
Never shared
Zero third-party access. Your genealogy stays private.
Delete anytime
You control your data. Delete your GEDCOM and all conversations with one click.
5. Multi-Layer Guardrails
Large language models love to fabricate. We prevent this through:
- ▸Hard cutoff enforcement — Ancestors cannot acknowledge events outside their lifetime
- ▸Fact verification — Every factual claim must cite a GEDCOM source
- ▸Conservative inference — Cultural context is allowed, but personal events cannot be invented
Our Named Concepts
We've developed specific terminology to describe our approach:
Provenance Markers
Citations like [P1], [P2] showing the source of every factual statement
Synthetic Genealogical Persona
Fictional but historically accurate ancestor built from constructed GEDCOM data
Blueprint-Based Reconstruction
Building AI personas from structured, verified-only genealogical profiles
Context-Aware Historical Grounding
Using era-appropriate cultural knowledge without inventing personal history
Why This Approach Matters
Genealogy isn't just data—it's memory, identity, and legacy.
When AI tools fabricate historical details or misrepresent ancestors, they don't just make technical errors—they distort family history and erode trust.
By building transparency, provenance, and ethical guardrails into every layer of ThroughTheirEyes.ai, we're setting a new standard for what AI genealogy can be:
- ✓Accurate — grounded in real genealogical records
- ✓Transparent — every claim cited and verifiable
- ✓Private — your family tree stays under your control
- ✓Respectful — treating ancestors as people, not data points
Who This Is For
ThroughTheirEyes.ai is built for people and organizations who take genealogy seriously—and we maintain clear rules on who can be recreated and why.
Genealogists & Family Historians
Researchers who need accurate, verifiable ancestor simulations
Family Historians
People preserving their family stories for future generations
Educators
Teachers using historical personas for immersive learning
Cultural Heritage Organizations
Museums, archives, and heritage groups preserving community history
Learn more about our ethical use framework including eligibility rules, data protections, and respectful storytelling guidelines.
Everything described on this page reflects features that exist in the platform today — not future promises. Our ethical, transparent, blueprint-driven approach is already live and shaping every conversation.
See It In Action
Talk to William Price and experience our approach firsthand.
Building an ancestor persona from your own GEDCOM? Join the beta.