PromptQL Audio Assistant

Hasura DDN Project available at: https://console.hasura.io/project/organic-liger-4989

This Hasura DDN supergraph enables you to query in-ear monitors (IEMs) and their frequency responses. The AI assistant enabled by Hasura PromptQL helps you explore the sound signature of a specific IEM and compare it with other IEMs or target responses. Also get the IEM recommendations based on your sound preferences and genres.

Available Target Responses

• JM1 (new meta)
• Harman
• Crinacle (IEF 2023)

Use Cases

Sound signature of an IEM

• The assistang would explain the sound signature and tonality of an IEM.

• Also provides comparisons with target (preference) curves

• Compare IEMs

• Get Eq recommendation

Get IEM recommendations

Getting started

Load measurements on your Postgres database.

Create the required tables:

-- Create the `iems` table to store information about each IEM
CREATE TABLE iems (
    iem_id SERIAL PRIMARY KEY,
    iem_name VARCHAR(255) UNIQUE NOT NULL
);

-- Create the `frequency_responses` table to store frequency and amplitude data for each IEM
CREATE TABLE frequency_responses (
    response_id SERIAL PRIMARY KEY,
    iem_id INT REFERENCES iems(iem_id) ON DELETE CASCADE,
    frequency_hz REAL NOT NULL,
    amplitude REAL NOT NULL
);

-- Create an index on the `frequency_hz` column for efficient querying
CREATE INDEX idx_frequency_responses_frequency ON frequency_responses (frequency_hz);
CREATE INDEX idx_frequency_responses_iem_id_frequency ON frequency_responses (iem_id, frequency_hz);

Load the measurement data in your database.

cd measurements
pip install -r requirements.txt

PG_URL= <your-postrgres-url-string> python load_db.py

Load eartip reviews on your Postgres database.

Create the required tables:

CREATE TABLE eartip_reviews (
    product TEXT PRIMARY KEY,
    bore_size TEXT NOT NULL,
    stem_length TEXT NOT NULL,
    feel TEXT NOT NULL,
    bass NUMERIC NOT NULL,
    midrange NUMERIC NOT NULL,
    treble NUMERIC NOT NULL,
    soundstage NUMERIC NOT NULL,
    vocal_presence NUMERIC NOT NULL
);
-- Indexes for optimizing queries
CREATE INDEX idx_product ON eartip_reviews (product);
CREATE INDEX idx_numeric_ratings ON eartip_reviews (bass, midrange, treble, soundstage, vocal_presence);

Load the eartip reviews data in your database.

cd eartips
pip3 install -r requirements.txt

PG_URL= <your-postrgres-url-string> python load_db.py eartip_reviews.txt

Set up Hasura DDN project

• Create a Hasura account if you don't have on already.
• Install Hasura DDN CLI.
• Login with CLI
• Set up .env file by cp .env_example .env
• Add Anthropic API key to .env. Get an api key from https://console.anthropic.com/settings/keys

# .env
...
ANTHROPIC_API_KEY=<your-anthropic-api-key>

To use an OpenAI key instead, you’ll have to set OPENAI_API_KEY in your .env file and change the environment variable LLM to openai in the compose.yaml file.

• Add your Postgres database url (must be accessible outside localhost) to the APP_PG_CONNECTOR_CONNECTION_URI in .env
• Setup your DDN project ddn project init
• Fire up your PromptQL project by
  • ddn supergraph build local
  • ddn run docker-start
• Open the PromptQL playground by ddn console --local

Now you’re ready to query the sound signature of your chosen IEM and compare it with others or with target responses.