The Simplest
(and fastest)
chatbot builder

Build a chatbot using the latest large language and ML models without drowning in microservice complexity.

try it for yourself!
Change the Brain:
Knowledge Base:
Clear History

Chatbot

How do I
use pgml.transform()?
Show me
a query to train a model
What is HNSW
indexing
Teach me
how to use pgml.embed()
What are
tensors?
How do I
train a model?
Show me
some features of PyTorch
Explain
how to use an optimizer?
What is
a lifetime?
How do I
use a for loop?
Show me
an example of using map
Explain
the borrow checker
How do I
join two tables?
What is
a GIN index?
When should I
use an outer join?
Explain
what relational data is

4x faster

than HuggingFace + Pinecone

26k + models

Llama, Falcon, Mistral, etc.

1m QPS

queries per second on EC2
"I have a full proof of concept chatbot fully synced
to document changes, all done in 3 hours flat."
Founder @

Keep it simple. Keep it fast. 

Today’s chatbot implementations require a patchwork build of services that introduce latency at each step. PostgresML combines and automates the entire chatbot workflow. It’s less infrastructure overhead and a better (faster) experience for users.

Old Way and New Way Old Way and New Way

4x Faster

than HuggingFace + Pinecone
for a RAG chatbot

10x faster

than OpenAI for embedding
generation

Save 42%

On vector database cost
compared to Pinecone

Take your
chatbot to
the next level

Take your chatbot to the next level

Deploy a next-gen chatbot with a cli builder, vector search, retrieval augmented generation (RAG) and the latest LLMs – all in your database.

deployed_code
Use a single database

Store your documents, chunks and text embeddings in one place for a simplified infrastructure footprint that requires less eng resources.

map
Generate text embeddings

Get the only platform that can generate state-of-the-art LLM models without an external LLM inference service.

inventory
Manage vector indices

Get lightning fast and accurate vector recall. Eliminate roundtrip network calls for recall and querying for the lowest latency app.

search
Monitor performance

Track chat history, prompts and prompt templates. Fine-tune the latest LLMs with chat history right in the database.

Chat wherever
your users are

Chat wherever your users are

Bring factual memory and lightning-speed responses to your website, Discord, Slack and more with a seamless integration to your preferred communication platform.

Build a chatbot as
easy as one, two, three

It’s simple with a seamless in-database MLOps platform.

Upload

1. Ingest

1. Ingest
pgml-chat --stage ingest
Upload

2. Verify

2. Verify
pgml-chat --stage chat \ --chat_interface cli
Upload

3. Deploy

3. Deploy
pgml-chat --stage chat \ --chat_interface slack

Start your free
project today

Sign up and start building for free today.
Our API and SDKs make it easy.

Start building with PostgresML

Start your free
project today

Sign up and start building for free today.
Our API and SDKs make it easy.

Start building with PostgresML