Ongoing Support
Regular engineering hours alongside your team.
MonthlyExplore
WhatsApp AI assistant
An LLM assistant that answers people on WhatsApp from the client's own content, running serverless on AWS, with a cost model to price it.
- Area
- Built with
- AWS LambdaBedrockS3 VectorsAurora Serverless v2React
- Related services
- Ongoing Support · Device Security Review
Benefits
What it delivers
01
Grounded answers
Replies come from the client's own content, not generic web text.
02
Pay per use
Serverless, so cost follows actual usage.
03
Priced with data
A cost simulator set the customer's pricing before launch.
How it works
From input to result
The path data takes through the system. The highlighted step is where the main work happens.
- 01WhatsApp
- 02Lambda
- 03Bedrock + RAG
- 04Aurora Serverless
- 05Reply
What I built
- Serverless backend on AWS Lambda and Aurora Serverless v2
- Retrieval-augmented answers with Bedrock and S3 Vectors
- React cost simulator covering model tokens, database, security services and WhatsApp per-message fees
Tech stack
AWS LambdaBedrockS3 VectorsAurora Serverless v2React
Services
Need something like this?
These services cover the same kind of work.
Device Security Review
Find the weak points before someone else does.
1–3 weeksExplore
Building something similar?
Tell me what you're working on. I'm glad to talk it through.