hanseong.kim Get in touch
Independent AI product studio

Tell it what to watch.
Get only what matters.

We're developing a web monitoring agent for booking availability, price changes, delivery updates and new posts. The workflow pairs plain-language rule drafting with human approval and scheduled checks, so you can spend less time refreshing pages.

Request early access → See how it works
Human-approvedAI drafts, you confirm
Rule-based checksNo model call on routine checks
Recoverable deliveryPersistent alert queue

The problem

The things people wait for still live on pages that don't notify anyone.

Cancellations, restocks, booking slots and tracking updates can change between visits. Checking them manually takes time, and useful changes can be hard to distinguish from page noise.

Challenges we're addressing

  • Separating meaningful changes from ads and layout updates
  • Setting up rules without writing selectors by hand
  • Distinguishing a failed check from an unchanged page
  • Recovering pending alerts after a restart
  • Keeping routine checks independent of model inference

Our approach

  • Alerts on the meaning — a new slot, a met price, a new step
  • Set up by describing it in a chat message
  • Only a fully parsed snapshot can replace the last good state
  • State change and alert are committed in one transaction
  • AI helps draft and revise rules; routine checks use plain code

How it works

AI where it's smart. Code where it must be exact.

The intended workflow uses a model to propose a rule, then hands routine checks to a deterministic engine. Drafts and later rule changes require your review.

Describe

Send a URL and what you want in a chat message. No selectors, no code.

“Alert me if this lens drops under 1.1M won”
LLM

Draft

An AI worker reads a sanitized copy of the public page and proposes a structured rule as strict JSON.

{"selector": ".price", "type": "number", "op": "<=", "value": 1100000}

Approve

You review the draft and approve, edit, retry or hand it off. Nothing goes live without you.

/approve 12

Watch

A sequential scheduler runs the rule on a fixed interval and alerts only on real transitions.

checks: 61s · AI calls: 0
Why this matters: routine checks do not need model inference. Drafting, retries and repairs can require additional model calls; site availability and check intervals affect when changes are detected.
Web monitoring architecture: a chat request goes to an AI draft worker, then to human approval. Approved rules run in a scheduler that polls target sites and commits state and queued alerts in one database transaction. A watchdog monitors the scheduler. Chat request URL + intent AI draft worker page → typed JSON rule Human approval approve · edit · hand off Target sites fixed interval, validated Scheduler parse → snapshot → diff State + outbox one database transaction Alerts sound / silent Watchdog heartbeat · restart

Reliability design

A monitor is only useful if you can stop thinking about it.

The architecture focuses on recovery, explicit failure states and reviewable rules. These mechanisms reduce operational risk; they do not guarantee uninterrupted monitoring or delivery.

Transactional outbox

State changes and pending alerts share a database transaction, allowing queued alerts to be retried after a restart. Delivery still depends on the receiving service, and retries can produce duplicates.

Last-known-good state

Snapshot validation is designed to preserve prior data when a response is empty, incomplete or cannot be parsed, helping distinguish outages from availability changes.

Cross-checked sources

Where supported, comparing two permitted views of the same data helps flag inconsistent availability before generating an alert.

Heartbeat & watchdog

Heartbeat checks and a separate watchdog help detect stalled monitoring and support recovery after a process failure.

Alerts that respect attention

Separate change alerts from routine status updates, with quieter recovery messages and controlled escalation for repeated errors.

Security boundaries

Design priorities include validating target URLs, restricting access to private networks and excluding secrets from logs. These controls need ongoing testing as the product develops.

Monitoring use cases

Appointments · availability changes Parcel tracking · auto-closes on delivery Room availability · supported booking pages Public announcements · new and updated posts Price & stock · numeric and text conditions Supported page regions · scoped change detection

The intended scope is public pages and APIs that permit automated access. Support depends on the source site's terms, access limits and technical behavior; private or restricted content is outside the intended scope.

Planned Claude integration

Better drafts, with a human in control.

A wrong rule either spams or stays silent — and users only notice the silence when it's too late. We need a model that reads messy real-world HTML carefully, follows a strict output schema, and says “I can't support this site” instead of guessing.

We plan to evaluate Claude for rule drafting and repair. The Claude API is not yet integrated. Draft quality will be measured against test cases before release, and users will review proposed changes.

01

Planned: intent → rule drafting

Use a request and sanitized page content to propose a structured monitoring rule, with supporting evidence and explicit uncertainty.

02

Planned: site support assessment

Assess whether a page or documented API can support the requested monitor. Unsupported cases should be flagged for review rather than guessed.

03

Planned: assisted rule repair

When a page redesign breaks a selector, propose a repaired rule and ask the user to confirm before applying it.

04

Planned: evaluation before release

Test draft accuracy, unsupported-site handling and repair proposals on representative pages before enabling the integration for users.

Selected work

Independent software projects.

Selected projects explore visual recognition, search, simulation and data visualization. They inform our approach to building and testing useful software.

Mobile · vision + searchIndependent project

Puzzle path planner

Explores number-tile board recognition and move planning, using visual previews to explain candidate moves while a background search improves the plan.

Visual recognitionSearch algorithmsMove previews
Desktop · scientific vizIndependent project

3D synoptic weather viewer

Explores surface pressure and upper-air weather-model data in an interactive 3D view, with pressure-center detection and vertical profiles for scientific visualization.

Layered weather dataVertical profiles3D visualization
Browser extension · searchIndependent project

Listing verification assistant

Explores cross-referencing public property listings, distinguishing direct matches from related or estimated results so a user can review the evidence.

Listing comparisonEvidence reviewGraded confidence
Mobile game · custom physicsIndependent project

Soft-body physics racer

A tilt-and-touch driving game built around a custom spring-mass simulation, with deformable vehicles, a speed-adaptive camera and interactive courses.

Custom physicsTouch controlsCourse design

How we build

Principles we don't compromise on.

We aim to keep design decisions and verification evidence close to the code as our projects develop.

Never guess silentlyUnknown is a state. We show it rather than fill it in.
Privacy by architectureProcess on-device where possible; collect nothing we don't need.
Verify on the real thingReal devices, real sites, recorded evidence — not just green unit tests.
Humans approveAI drafts and explains. People decide what runs.

Roadmap

From a personal tool to a hosted service.

The current project is a self-hosted agent. Claude API integration and a hosted multi-user service are planned next steps, not currently available features.

Now

Self-hosted agent

  • Chat-bot control & natural-language drafts
  • Six monitor types, generic page + condition rules
  • Outbox, watchdog, heartbeat, error escalation
Next

Claude-powered drafting

  • Claude API with tool use for typed rule drafts
  • User-reviewed selector repairs after site redesigns
  • Eval suite of real pages for draft accuracy
Later

Hosted multi-user beta

  • Per-user isolation, quotas and billing
  • Web dashboard alongside chat control
  • Templates for supported monitoring use cases

Contact

What would you like to monitor?

Interested in early access or sharing a monitoring use case? Tell us what you'd want monitored. The hosted service is still in development.

Request early access → General inquiry
admin@hanseong.kim