
eFisher
eFisher is a phone app for fishers, boat owners and fish traders on Lake Victoria, in Kenya, Tanzania and Uganda. It keeps trip and boat accounts, runs a small fish market, sends tomorrow's weather every afternoon, and answers questions from fisheries documents. It was built for the Lake Victoria Fisheries Organization and runs on the FAIMS platform.
- registered users
- 234
- countries
- 3
- fish posted for sale
- 112
- boats on record
- 112
Data from the people who fish
The catch survey sees a boat once, on the day an enumerator happens to sample it. A boat owner sees every trip. eFisher gives owners, crews and traders a reason to keep their own records, a clear view of what each trip earned and cost, and in return builds a picture of catches, prices and effort that no survey could afford to collect.
Users sign up with a phone number and a one-time code, and choose their country and language. Species have their local trade names, such as omena, dagaa and mukene for the same small sardine, so a trader in Kisumu and one in Mwanza mean the same fish.
Trips and boats
An owner registers boats and assigns crew. For each trip the app records departure and return, the district and landing site, the catch by species with count, weight and price, and every expense, fuel, repairs, food, crew allowances and the rest, each with its unit and unit cost. From that it draws the owner’s income, spending by category and the trend over time. Traders get the same view of what they bought and sold.
Fish market
Anyone can post fish for sale: species, condition, selling unit, price and currency, place and a photo. Posts are checked by an administrator before they go live. Approved posts can also be published to the eFisher Facebook page and Instagram account, with each attempt logged and retried, so buyers who never install the app still see what is on offer. Sellers mark a post sold, in full or in part, and the price they actually got.
Weather on the lake
Users save the places they fish. The app shows the forecast for each, and for points on the lake adds a one-line note on whether the conditions look safe and worth going out, written by a small language model from the forecast figures.
Every day at 16:00 East Africa time each user gets tomorrow’s forecast for their saved places by SMS, and by WhatsApp where they use it. A text message reaches a basic phone at a landing site with no data bundle, and on Lake Victoria storms can build within hours.
Ask a document
Fisheries guidance tends to sit in long PDFs nobody reads on a phone. In eFisher the user picks a document and asks a question in plain words. The whole document goes to a language model, Gemini, with the instruction to answer only from it. Two documents are loaded now: a market survey of Lake Victoria’s small fish trade, and the guidelines for the East African standards on dagaa, omena and mukene.
Users can flag an answer and say what was wrong with it, and those reports are kept so the documents and instructions can be improved.
On the web platform
Fisheries staff manage eFisher from FAIMS: users and their boats, posts waiting for approval, trips, analytics on users, posts and catches by species and month, and logs of every sign-in and change. Phone numbers and names of users never leave that view.
Results
eFisher is still a pilot and the numbers are small: too small to say anything about catches or prices on the lake, and not shown here for that reason. Figures are refreshed from FAIMS every night. They are provisional and not for citation.