Zero to Operational for Amazon Neptune using Agentic AI
This series of articles will show how you can go from just a concept for a graph database powered workload to fully operational using Agentic AI to lead the way
UPDATED 2025-Dec-3: Added links to the Query Tuner agent, completing the series.
UPDATED 2025-Nov-21: Added links to the Sample Data Generator, Query Generation, and Scenario Data Generator agents.
UPDATED 2025-Nov-10: Added links to the Use Case Generator and Modeler agents.
This is the start of a new series of articles where I will share my ongoing research on enabling customers to go from concept to fully operational solution on Neptune. It is going to be long running and may change as my research, results, and the technology stacks evolve, but stay tuned here to follow in the journey. Follow me on LinkedIn if you want to know when I add to this, as I will post updates there.
Here is the journey we will take:
- Use Case Generator Agent - start with a concept and build a use case for your graph workload
- Modeler Agent - tell it your use case and get a graph model
- Sample Data Generator - give it your model and these agents will populate your Neptune database with realistic data
- Query Generation - tell it what questions do you want to ask of your graph and it will look at your graph model, generate the queries (testing them to make sure they do work or refining them if they do not)
- Scenario Data Generator - did the sample data generator not generate the scenarios that your complex questions were designed to detect (e.g., money laundering patterns for fraud detection)? Give this agent your query and use case and it will generate complex scenarios for you.
- Query Tuner - give it your query and it will run a Neptune explain plan to analyze any inefficiencies, generate some alternative queries and test them for speed and whether they give the same results before proposing them to you.
Other areas we might expand into down the road:
- Automatic ETL
- Observability Agents watching your cluster for you
- Explain this query profile in plain English (this one is closer than you might think)
Anything you want to see that I don't have covered here? Let me know.
Enjoyed reading this content? Let the author know!
Your likes, comments, shares, and saves help creators reach more builders.
Loading recommendations
Loading article