The Vision
We are not building an immigration app.
We are building a platform for navigating complex institutional systems.
We structure user data, apply rule-based and AI logic against live policy sources, and generate dynamic action plans that evolve over time. Immigration is where we start - because it is the hardest regulatory journey a person can face. The system remains identical across industries. Only the policy data changes.
How It Works
One engine. Any regulated domain.
Policy Data
Any regulatory source. Government rules, institutional processes, compliance frameworks.
Settlewell Engine
Scores eligibility. Maps pathways. Generates plans. Monitors changes. Adapts in real time.
Identical across all domains
Live Action Plan
Personalized, multi-step roadmap that updates as conditions change and progress evolves.
Unique to every user
Why immigration first
The hardest version of the problem.
If we can build a system that guides someone through 100+ immigration pathways, constantly changing government policy, multi-year document timelines, language barriers, and post-landing settlement — the same infrastructure guides anyone through any complex regulatory system. Immigration is our proof case. The platform is the product.
100+ interconnected programs
Federal, provincial, and territorial pathways with overlapping requirements, shared eligibility rules, and complex dependencies.
Policy changes continuously
IRCC updates requirements, draw patterns, and program rules regularly — any guidance that's one month old can be wrong.
Journeys span years, not days
From first eligibility check to citizenship can take 5-10 years — requiring a system that persists, adapts, and remembers.
Mistakes have real consequences
Wrong pathway selection, missed deadlines, or incorrect form responses can mean years of delay or permanent ineligibility.
Setti - The AI Layer
A domain-agnostic AI engine.
Fed any source, trained on any domain.
Not a general-purpose chatbot. A specialized engine that ingests verified data sources, operates on a user's real profile, and escalates to professionals when confidence is low. Immigration is the proof case. The architecture applies to any regulated domain.
Setti
Online - grounded in official sources
Verified-source citations
Only answers from authorized sources. IRCC, health authorities, CRA - you control the data.
Full profile awareness
Every answer is personalized to the user's exact situation, scores, and stage.
Uncertainty gating
Routes to qualified professionals when confidence is low. Does not guess.
90+ languages
Domain-specific terminology preserved across every language.
The Stack
Six technology pillars. One seamless system.
Every layer was built from the ground up for the complexity of long-term regulatory navigation. Here is what powers the system.
An always-current map of every rule
A system that continuously reads official government sources, structures them into a knowledge graph, detects when rules change, and determines how each change affects each individual user.
- Structured understanding of 100+ programs and their requirements
- Automatic policy change detection with personalized user impact
- Full version history for reproducible, auditable guidance
- Cross-program dependency mapping
A personalized navigation system for every person
A dynamic planning engine that generates a step-by-step, dependency-aware roadmap unique to each person, automatically replanning when circumstances change, and always surfacing the highest-value next action.
- Roadmap with task dependencies and parallel tracks
- Automatic replanning when profile or policy changes
- Multi-dimensional scoring: eligibility, readiness, risk, fit
- Scenario simulation for profile improvements
AI grounded in truth, not probability
An AI guidance layer that assembles responses from a user's real profile, live policy data, and conversation history — citing only official sources, with structured multi-step reasoning for complex questions.
- Context-aware responses grounded in real-time official sources
- Structured multi-step reasoning for complex queries
- Confidence-gating and professional escalation routing
- Full advice reproducibility for compliance
Intelligence that never stores what it shouldn't
A privacy-first layer that extracts structured profile data from natural conversations using a strict allow-list schema, automatically detects and redacts sensitive identifiers, and flags inconsistencies across all data sources.
- Profile extraction from conversation with privacy-by-design constraints
- Real-time detection and redaction of sensitive identifiers
- Cross-source consistency validation
- PIPEDA and Quebec Law 25 compliant, data stored in Canada
Guidance that continues after approval
A post-approval intelligence layer that scores settlement readiness, recommends optimal cities based on individual profiles, and connects users with the right services at the right moment in their lifecycle.
- Multi-dimensional settlement readiness scoring
- City recommendation engine matching profiles to regional outcomes
- Journey-timed partner and service matching
- Lifecycle tracking from PR through citizenship eligibility
Anticipating what comes next
A predictive layer that models processing outcomes, analyzes patterns in institutional decision-making, benchmarks users against cohort data, and surfaces insight ahead of time — not after the fact.
- Case-specific processing time prediction based on profile similarity
- Decision pattern analysis to optimize submission timing
- Anonymized cohort benchmarking against similar outcomes
- Profile strength scoring with highest-impact improvement recommendations
Platform Vision
Built for immigration. Designed for everywhere.
The core system ingests regulatory data, maps it to a user's profile, scores eligibility across pathways, and generates a personalized multi-step plan. It monitors for policy changes and adapts in real time. This is not a product that needs to be rebuilt for each industry. It is an engine. Immigration is where we prove it works.
We are building the method - not the market.
Every patent family we are filing covers the system architecture, not the vertical. The claims apply regardless of which regulatory domain the engine operates in.
Global Immigration
Same engine applied to US, UK, Australia, EU immigration systems. Different policy data, identical architecture.
Corporate Compliance
Same scoring and tracking engine applied to corporate visa management and workforce authorization.
Healthcare Navigation
Same decision engine applied to insurance systems, treatment pathways, and benefits navigation.
Tax and Benefits
Same rules engine applied to multi-jurisdiction tax codes and government benefit programs.
Education Pathways
Same journey planner applied to credential recognition, admissions, and academic transitions.
Government Services
Same platform applied to any citizen-facing institutional process requiring structured, ongoing guidance.
Six of the regulated domains where the same engine applies. There are dozens more.
Intellectual Property
Built to be defensible from day one.
Settlewell's technology strategy was designed alongside its product strategy. Every major capability is either in active patent filing or protected as a trade secret. The result is a platform architecture that is genuinely difficult to replicate — not just a collection of AI prompts on top of an existing model.
Our IP portfolio covers the full stack: how we compile and version government policy, how we generate personalized roadmaps, how our AI reasons about uncertainty in regulated contexts, and how we maintain full advice reproducibility for compliance purposes.
8
Patent families spanning policy intelligence, journey planning, AI orchestration, privacy, settlement, and predictive analytics
42
Features identified for patent filing. Provisional applications in preparation, pending funding
22
Trade secret features protected through proprietary implementation, access controls, and first-mover advantage
48+
Features with no direct equivalent in any Canadian consumer immigration product today
Competitive Landscape
What changes.
| Capability | Traditional Consulting | Static Information | General AI | Settlewell |
|---|---|---|---|---|
| Personalized step-by-step roadmap | Partial | No | No | Yes |
| Real-time policy change detection | No | No | No | Yes |
| Personalized policy impact analysis | No | No | No | Yes |
| Multi-dimensional readiness scoring | Partial | No | No | Yes |
| AI grounded in official citations only | No | No | No | Yes |
| Scenario simulation for profile improvements | No | No | No | Yes |
| Post-landing settlement guidance | Partial | No | No | Yes |
| Processing time prediction (ML-based) | Estimate only | No | No | Yes |
| Privacy-first profile extraction | No | No | No | Yes |
| Available 24/7 in 90+ languages | No | Partial | Partial | Yes |
Personalized step-by-step roadmap
Real-time policy change detection
Personalized policy impact analysis
Multi-dimensional readiness scoring
AI grounded in official citations only
Scenario simulation for profile improvements
Post-landing settlement guidance
Processing time prediction (ML-based)
Privacy-first profile extraction
Available 24/7 in 90+ languages
See the platform in action.
Explore the platform preview to see what Settlewell looks like from a user's perspective. Full backend in active development.