🎙️ Transcription Distiller

SLM Meeting Summarizer

Offline transcription post-processor. Distills meeting transcripts into action trackers, schedules, and bulleted logs with strict formatting rules.

🚀 Overview & Capabilities

Offline transcription post-processor. Distills meeting transcripts into action trackers, schedules, and bulleted logs with strict formatting rules.

Key Features

Turns conversational text blocks into formal action tables
Identifies speaker intent, decisions, and deadlines
Map-Reduce pipeline support for 2-hour long transcription logs
Strict template outputs matching markdown specifications

💻 Installation

Install the local CPU-optimized package using pip:

Terminal

              # Install local CPU-optimized package
pip install slm-meeting
            

🐙 Checkout from GitHub

Clone only this agent's folder from the monorepo using Git sparse-checkout — no need to download the full repository:

Option 1 — Sparse Checkout (Recommended)

Terminal — Git Sparse Checkout

              # 1. Create and enter a new directory
$ mkdir slm_meeting && cd slm_meeting

# 2. Initialise empty git repo and add remote
$ git init
$ git remote add origin https://github.com/t00114218-stack/SLMAgents.git

# 3. Enable sparse-checkout and set target folder
$ git sparse-checkout init --cone
$ git sparse-checkout set slm_meeting

# 4. Pull only that agent's source
$ git pull origin main
            

Option 2 — Full Repository Clone

Terminal — Full Clone

              $ git clone https://github.com/t00114218-stack/SLMAgents.git
$ cd SLMAgents/slm_meeting
            

💡 Tip: After checkout, install the package locally with pip install -e ./slm_meeting to run in editable mode without publishing to PyPI.

⚙️ Configuration API

Constructor Parameters

Instantiate SLMMeetingSummarizer with performance options:

Parameter	Type / Default	Description
model_path	str \| None	Explicit path to ONNX model weights. If omitted, downloads standard checkpoints.
cache_dir	str \| None	Directory to store model weights offline. Defaults to `~/.cache/slm-meeting/`. Also settable via SLM_MEETING_SUMMARIZER_CACHE_DIR.
n_threads	int \| 4	CPU thread count for ONNX inference. Optimize for CPU core count. Also settable via SLM_MEETING_SUMMARIZER_N_THREADS.

Methods

Method Signature	Return Type	Description
`summarize_transcript(transcript_text, system_prompt=None, user_input=None)`	`dict`	Groups speakers, extracts decisions, and renders action items into strict Markdown tables.

Method Parameters (Execution Customization)

All main execution methods accept optional system routing parameters:

Parameter	Type / Default	Description
system_prompt	str \| None	Optional custom system prompt instruction to override the default system template response parameters.
user_input	str \| None	Optional additional user-supplied target text variables or contextual keys.

Quick Start

from slm_meeting import SLMMeetingSummarizer

summarizer = SLMMeetingSummarizer()
todos = summarizer.summarize_transcript(
    transcript_text,
    system_prompt="Extract action items only",
    user_input="Deadline format: YYYY-MM-DD"
)
print(todos)

Environment Variables

Configure agent parameters globally using environment values:

Environment Variable	Default	Purpose
SLM_MEETING_SUMMARIZER_N_THREADS	4	Sets CPU inference execution threads.
SLM_MEETING_SUMMARIZER_CACHE_DIR	~/.cache/slm-meeting/	Default directory to store downloaded ONNX weights.

CPU Performance Tuning

To run the SLMMeetingSummarizer engine efficiently on CPU under 1.5 GB memory footprint:

Match Threads to Core Count: Set n_threads or SLM_MEETING_SUMMARIZER_N_THREADS to match the physical CPU core count.
Sequential Processing: Avoid concurrent processing when batch files are large.
Garbage Collection: Clear variables and run gc.collect() to release model RAM blocks after execution.

Verified Input & Output Logs

Diagnostic execution console response running locally on CPU:

→ INPUT:
"Alice: I will deploy the schema."

← OUTPUT:
{
  'speakers': ['Alice'],
  'action_table': '| Speaker | Assigned Action Item | Deadline |\n| Alice | I will deploy the schema. | TBD |'
}