Anyone else maintaining a wildlife species database solo? How do you keep it accurate?

General
1

Hi everyone,

My name is Tianhao Peng, and I’m an independent developer working on safarilingo.com — a wildlife animal guide app build for tourists visiting Africa.

I’ve been building it mostly on my own, with the hope of making it easier for travelers, guides, and wildlife lovers to learn about what they’re seeing in the field. The app currently includes identification info for a few hundred African animals, along with sounds and sighting records shown on a GPS map.

I’m posting here mainly because I could really use some advice.

One challenge is that maintaining an animal database alone is honestly pretty hard. Once you start covering hundreds of species, it becomes stressful to make sure the information stays accurate, balanced, and updated. Right now, a lot of my corrections and additions come from user feedback, which has been helpful, but it’s still a bit fragile as a long-term quality control system.

So I wanted to ask: for those of you who work with species data, field guides, or community science projects, are there any good ways to handle quality control as a solo builder? And would there be any better structure for community review? Even lightweight ideas would help.

The second reason I’m posting is to say thank you. Some of the animal sound data in SafariLingo comes from the iNaturalist community, and I genuinely appreciate that this ecosystem exists. It has made the app much richer than I could have done alone.

At the same time, I’m not sure how best to give back.

The app has started to accumulate some wildlife sighting records from users traveling in African savannas, including GPS-linked observations and photos. I feel like some of that could potentially be useful to the iNaturalist community, but I don’t know what the right path is. Is there a recommended way to contribute data back? Would that mean encouraging users to post directly to iNaturalist, exploring some kind of export format, or something else entirely?

I’m not here to advertise — honestly, not that many people go on African safaris anyway. I’m trying to build something carefully, learn from people who know this space much better than I do, and hopefully contribute in a respectful way over time.

Thanks for reading, and I’d really appreciate any suggestions.

2

I simply select the area I would like covered; in my case it is

https://www.inaturalist.org/observations?place_id=130719

and let the user select what s/he wants to know about it ;-)
and, Yes! I encourage people to open a (free) username and contribute.

3

I’m not sure if I’m misunderstanding what you’re aiming for, but it seems like you’re doing a lot of work to duplicate something that already exists.

Maybe you should consider:

  • An iNat Project (possibly a trafitional one, so you can add just the species you want)
  • A document with your written summary of each species but a link to specific iNat observations for more details.
4

For birds I import AviList into my private database with appropriate code to work out what has changed with a manual review process - for a long time based upon IOC and very sad to see it’s demise. For everything else variously manually creating and maintaining taxa and worst case I just create anything I don’t have as and when I see it or for Singapore as and when anyone I trust sees something. As said by someone else it’s a lot of work - my personal taxonomy currently has 60,995 nodes!

5

to track inat taxonomy changes of my sightings I use this query https://api.inaturalist.org/v1/identifications?current=true&is_change=true&own_observation=true&user_id=mike_hooper&year=2026&per_page=30&page=1

6

Their app looks to be quite different from iNaturalist.

7

Thank you so much for your reply! The biggest difference is that the core purpose of this app is for users to tag animals on grasslands without internet access, and then learn about the animals. Essentially, it serves the tourist’s travel experience and their memories after returning home (although some NGOs have recently sought cooperation, they are not targeting the core users).

Therefore, any way to obtain more accurate data would be extremely helpful.

8

With all due respect, a good first step would be to make sure your AI-generated images actually look like the species they purport to show. A good number of the birds are totally wrong.
Or, better yet, use some of the many Creative Commons images in the iNat database instead.

9

Yes, that’s why I want to solve this. Maintaining a professional image library isn’t easy right now. Regarding open-source solutions: are the images on iNaturist currently licensed for commercial use?