The Fish-Assist

Lake plans and lure matches.

Pick a lake and a date, and get an ordered plan instead of a feed of tips: depth ranges, presentation, and lure matches ranked against the conditions and drawn only from tackle you actually own. Every recommendation shows the rule that produced it.

CatalogueDSL 004
Status In development
PlatformMobile · Web
Privacy Policy →

Fishing advice is abundant and almost entirely context-free.

A tip that worked on a different lake, in a different season, at a different water temperature, using a lure that is not in your box is not advice; it is noise. The thing that would actually change the day is the intersection of the water, the conditions, and what is in the boat with you, and very little puts those three together.

A fishing assistant that turns lake conditions and the gear you already own into a practical plan: an ordered set of things to try rather than a feed of tips.

Who it is forAnglers who fish the same lakes often enough to want a system, not a tip sheet. It is fully usable without an account.

In active development

How a plan gets made.

The Fish-Assist has no public build and no screenshots to show yet. What follows is the part that is settled: how a plan is produced, and what it is allowed to say.

  1. Lake The waterbody, or an explicit not-known if there is no cached record for it.
  2. Conditions Weather and water conditions for the chosen date.
  3. Season Where that date sits in the year for this water.
  4. Species What the conditions imply about depth, position, and behaviour.
  5. Owned tackle Your rods, reels, lines, and lures. Matches are drawn only from these.
  6. Plan An ordered set of things to try, each carrying the rule that produced it.
  • A recommendation that cannot name its rule is not shown.
  • A lure that is not in your inventory is never suggested.
  • An unknown waterbody produces a stated not-known, never a generic plan.
  • Plans generate offline from cached data.

Recommendation engine and tackle inventory

  1. Every recommendation shows its rule

    The rule that produced a suggestion, and where that rule came from, travel with the suggestion.

  2. Only gear you actually own

    Lure matches are drawn from your inventory. The app does not recommend a purchase in place of an answer.

  3. Plans generate offline

    Cached waterbody data means a plan can be produced at the ramp, not only at the desk.

  4. It says when it does not know

    An unfamiliar waterbody produces an explicit not-known rather than a plausible generic plan.

  5. Usable without an account

    The core planning loop does not require signing up for anything.

The technical surface, without the implementation
  • Condition-based recommendation engine
  • Tackle inventory matching against ranked presentations
  • Cached waterbody data for offline planning
  • Rule provenance carried on every recommendation
  • Multi-surface clients over shared recommendation services
Status
In development
Platform
Mobile · Web
Price
Free when closed testing opens

No public build yet. Testing opens when the recommendation engine is finished.