Future of Dgraph

Hey Dgraph Team,

Is there a timeframe on when you guys are going to make any changes? Better yet is there a timeframe on when you guys will let us know the time frame for anything about anything?

In case you guys don’t know, we rarely hear updates from 3 different teams of Dgraph teams, which is by far the greatest downfall of Dgraph. Crickets will kill a product faster than a bad product.

Please don’t tell us to stay tight, we have been doing this for years now.




I echo these sentiments. I’ve posted my thoughts on how Hypermode can lean into core Dgraph development on the semi-recent discuss topic: How are we tracking on our priorties?. I’d love to hear what the future has in store for DGraph.

I am also hesitating on leaning on Dgraph for new client projects until I have a bit of clarity/reassurance of the future of the product

Hi All-- this is my fault. With the transition, I’ve been heads down on integrating the teams and driving Hypermode’s products toward GA. That’s not been fair to you all. I’ll try to be much more active here going forward.

On the radio silence, the short answer is:

We’ve been working with Hypermode’s first few customers to implement our offering end to end with them. The first cohort will be going into production over the new few weeks.

Along the way, we’re documenting/implementing improvements to Dgraph including some of the showstoppers you all have shared (reliability, consistent performance, upgrades, better debugging/monitoring). We’re also adding quality of life improvements (Badger, GraphQL improvements), and new features (vectors). As promised, these improvements will shortly be rolled into Dgraph as open source improvements.

We’ve been largely silent because I’m sensitive to making a commitment to ya’ll then walking it back as priorities get shifted to support our commercial customers. My DMs are always open if you want to ping me specifically about an item.

Dgraph is still a vast majority of our revenue-- we’ve got several engineers still full time on improving it. We’re also still driving toward getting Dgraph into a place where it’s able to be managed/contributed to by the community/users rather than purely a commercial entity (Dgraph Labs).

We’ll have a clear “go forward” roadmap, community plan, release cadence, etc. by no later than the end of Q1 this year.


Can you share these?

I’m looking forward to this.

This is the man who decided to snap up the product. In my experience, the bigger the smile, the bigger the catch.
In my opinion, it was wrong to make the base in golang, and the requirements for the product are too great. I have been using surreal db with sql-like syntax for a long time. Or as an alternative you can look at memgraph. Though of course I liked the systaxis of dgraph better. And so for the community it is already a dead base without support and updates.

End of Q1 is approaching! Are you still expecting to have a public update ready by then?


I saw a bunch of commits and some note somewhere about developing in a private repo. That explains why the commits are so quiet. Hmm…

  • 92cf054 feat(dql) add similarity score & cosine similarity workflows (#8)
  • b7bf36a Feat(Indexing): Added delete support for vector indexes. (#12)
  • 0cdc4fa Sunilmujumdar/dgr 61 build hnsw index with existing nodes (#13)
  • 995080d feat(indexing): Move HNSW to private repo for development (#14)
  • b1de366 chore: allow workflows to checkout other private repos (#16)
  • 5812708 fix(dql): Repair a broken type assertion error. (#15)
  • dfd847e feat(indexing): Add Tokenizers for vfloat type (#17)

Long-time Dgraph user, also awaiting this end of Q1 update to help decide whether to continue that path or not.


@KVG @Dgraph_Admin - tagging here to make sure the message isn’t missed.

Times really counting down, unless your Q1 is longer than the first 3 months or maybe you don’t go by the physical month for quarters… :face_with_peeking_eye:

1 Like


Hi all, here’s my Q1 update. No April Fools jokes! :smiley:

1 Like