Help me decide between Dgraph and Neo4J (paid)

Users Dgraph
1

I’m looking to pay someone to help me decide between the two.

  • I’m gonna be more or less married to the one I choose, for the foreseeable future.
  • I haven’t touched a graph db ever before.
  • The graph db in the system I’m building is gonna be the primary source of truth.
  • Horizontal scaling is necessity, as this’ll be an OLTP SaaS
  • Application layer is running on Amazon Lambda

I’ve already posted a job on Upwork: https://www.upwork.com/jobs/~0176faa61b9c8fb076
However, none of the applicants has any experience with Dgraph. My hope is to talk with someone who has touched both DBs.

Feel free to ping me either on Upwork or here.

Regards!

1 Like
2

You might also want to look at Apache AGS which runs as an extension on PostgreSQL.
Dgraph still has a lot of potential but it’s unclear what it’s future is.

3

Hi Daniel,
If you are still on the lookout to for someone, I would be glad to bounce of some ideas as I am an experienced Dgraph developer.
You can reach out to me on my email here
Colin

4

Thank you, Colin, and everyone else who reached out!
After a couple of conversations, I realized graph DBs are a bite I’m not able to swallow just yet.

Things that stood out were managing infrastructure, as well as data modeling considerations and super nodes.

For the MVP of my system, I decided to store the graphs in S3, of all things : )

If god forbid, it gets traction, I’ll be back here looking out for experienced consultants.

Regards everyone!

5

DGraph: better DX, faster, less certain future, slow progress
Neo4J: established and mature product, super complex, can be slow

1 Like
6

We choose Dgraph over Neo4J because Dgraph is horizontally scalable. I believe it is the only open source graph database that is horizontally scalable. At the time we performed our evaluation (2020-ish), Neo4J was only vertically scalable.

1 Like
7

Daniel,

If you’re considering a paid version, let us know and we will assign some solutions architecture hours to answering your questions, or demonstrating how to do anything you need proof points on. We are biased in favor of Dgraph, but also don’t want to waste your time or ours by having you use Dgraph if it is the wrong fit.

I don’t think you’ll be as locked into whatever solution you start with as you think. LLMs can translate queries and schemas from tech to tech pretty well now, and it will only get better. Also GraphQL is a standard, so you can use that in many cases and avoid most of the tech lock-in.

Damon

1 Like
8

Thank you Damon!
I have already settled on an MVP architecture. If the project gets traction, I’ll be referring back here for sure!

FYI, my project is not related to language or any other models.

1 Like