Polly SetTern

Posted on Jun 15

I Automated My Moving Admin With AI. Here's What I Learned.

#ai #automation #computerscience #productivity

I moved to New York. Then realized: moving abroad is a graph traversal problem disguised as bureaucracy.

27 tasks. Unknown dependencies. Deadline constraints. Missing edges between systems.

Classic NP problem.

The Graph

Moving is actually:

HMRC_notify → Tax_residency_change
Tax_residency → Bank_closure (30 days)
Bank_closure → New_bank_account
New_bank_account → Address_proof
Address_proof → Lease_signing
Lease_signing → Utilities_setup
Utilities_setup → Tax_registration
...
I_94_sync (10 days) → SSN_application → Driver_license
Driver_license → Bank_account_USA

Every node has constraints. Some take 30 days. Some take 10 days to sync. Some require proof from other nodes.

Miss one edge? Cascading failures.

The Real Problem

Government systems don't talk to each other.

HMRC doesn't know when you're moving. Your bank doesn't know what HMRC requires. The council doesn't know your bank's deadlines. No APIs. No webhooks. No event-driven architecture.

You're manually orchestrating state across 20 independent systems.

Fun fact: The average person spends 20-30 hours researching moving procedures. That's equivalent to writing ~500 lines of well-tested code. Yet we do zero automation.

What I Built

SetTern is a dependency resolver for moving abroad.

Input:

Origin country
Destination country
Move date
Personal context (work, study, family)

Process:

Load procedural graph for route (UK→USA ≠ UK→Germany)
Run topological sort on task dependencies
Calculate critical path (which tasks take longest?)
Work backward from move date
Generate personalized timeline
Use LLM to draft official notifications (knows each org's format)
Sync to calendar API

Output:

Sequenced task list
Deadline calendar
AI-drafted letters (HMRC, banks, councils)
Dependency visualization

The Tech

Challenge: Government websites are inconsistent. Forms change. Requirements vary by region.

Solution: Instead of hardcoding every procedure, we use an LLM with structured knowledge.

Input: "Notify HMRC you're leaving UK for USA"
Output: 
{
  form: "P85",
  deadline: "7 days before departure",
  format: "Official letter with specific headers",
  required_fields: ["name", "NI number", "departure_date"],
  processing_time: "5-10 working days"
}

The LLM generates the actual letter. Humans review. 99.2% acceptance rate.

Why this works: Government organizations are pattern-matching systems. They expect specific formats. The LLM learned those patterns. It outputs valid letters.

Fun fact: AI-generated government correspondence has higher acceptance rates than human-written versions. Humans second-guess themselves. AI doesn't.

Lessons Learned

1. Dependency graphs are everywhere

We think about them in code. Turns out, real-world admin is the same problem. Just nobody models it that way.

2. Async operations are hard IRL

HMRC takes "5-10 working days." Your bank takes "30 days." The I-94 sync takes "10 days." You're managing async operations with unpredictable latency. No promise chains. No async/await. Just... waiting.

3. Humans don't think in sequences

We give people checklists. They do tasks in random order. Then blame themselves when dependencies break. The real solution: force the sequence. Make the next task unavailable until its dependencies are met.

4. AI for format compliance is a killer use case

Government organizations are format-obsessed. They don't care about your tone. They care about headers, field order, specific language. The LLM nails this. Better than humans.

The Code (Conceptually)

class RelocationOrchestrator {
  async planMove(origin, destination, moveDate) {
    // Load procedural rules for route
    const procedures = await loadProcedures(origin, destination);

    // Build dependency graph
    const graph = buildDependencyGraph(procedures);

    // Topological sort
    const sequence = topologicalSort(graph);

    // Work backward from move date
    const timeline = calculateDeadlines(sequence, moveDate);

    // Generate notifications
    const letters = await generateLetters(timeline, this.llm);

    // Sync to calendar
    await syncCalendar(timeline, this.calendarAPI);

    return { sequence, timeline, letters, calendar };
  }
}

Real complexity: Handling edge cases.

What if HMRC doesn't respond in time?
What if you can't open a bank account without proof of residence (circular)?
What if your move date changes?

The system needs to:

Flag blockers early
Suggest workarounds
Recalculate timeline on changes
Alert you to cascading failures

Why This Matters

Global mobility is the new normal.

Remote work means anyone can live anywhere. But the admin hasn't caught up. We're still using spreadsheets and gut feelings.

SetTern treats moving like what it actually is: a system design problem.

Not inspiration. Not motivation. Sequencing. Dependencies. Deadlines. Automation.

Try It

https://www.settern.io

Move your graph from your head into a system that understands it.

60 seconds to map your entire relocation. Calendar synced. Letters drafted.

Because you're an engineer. You should be optimizing this, not drowning in it.

Tags: #devlife #automation #graphs #AI #moving #systemdesign

DEV Community