summarize-meeting 0.1.1 → 0.2.0

checksums.yaml

@@ -1,7 +1,7 @@
 ---
 SHA256:
-  metadata.gz: d69584205af88a85d7c42bdcf5964d7e284523eaebc8a851245c6d807af1bc79
-  data.tar.gz: 6d581baaf28e859a2d78475076c6b131ce25678c56ffa73752803e80eb016075
+  metadata.gz: 3c14fc93afabfd456a2c3778dffccf9d2a66e77f2d47ab98b864986ad0e10cd2
+  data.tar.gz: 68ba17378b48f84ad029ef1d3d6e1575110739bb4c8f5c40e25a759350722cba
 SHA512:
-  metadata.gz: 8c0eb3b83b0b4f2503592c7ed954300622572ce0357dca6c24bb022ee3b6fc84d96647b5cc544732a56628d206f36502e003b7fab4498dd2c92a6e6d4059ab0f
-  data.tar.gz: 896c4cca235e708b43753edb078eb717cdba0580826a0d5b35382e71d5f59fe70121dedf362035040bbd216bcfc59d764bb3a56ccd8bfe70a9c9bce3456b375a
+  metadata.gz: 40a944d8e2289d02368c1f05c98ef3b26612f40c01c0aa761c110e7f46664b8411f11c7881152b32cc4aa3d1c4db20d726906fdc00ef69c5d362779cd5300e40
+  data.tar.gz: 6477d716ebf27e2b6b7a8a5e15c4ce2ec205e7f089b277002fb5f1ab5273d1e486f2248625571038475fac55e91d81fa77378094615727d0bccae970251b1f74

data/LICENSE.txt

@@ -0,0 +1,8 @@
+Copyright 2023 XBE LLC
+
+Permission is hereby granted, free of charge, to any person obtaining a copy of this software and associated documentation files (the “Software”), to deal in the Software without restriction, including without limitation the rights to use, copy, modify, merge, publish, distribute, sublicense, and/or sell copies of the Software, and to permit persons to whom the Software is furnished to do so, subject to the following conditions:
+
+The above copyright notice and this permission notice shall be included in all copies or substantial portions of the Software.
+
+THE SOFTWARE IS PROVIDED “AS IS”, WITHOUT WARRANTY OF ANY KIND, EXPRESS OR IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY, FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN THE SOFTWARE.
+

data/README.md

@@ -0,0 +1,96 @@
+## Summarize Meeting
+
+Automatically generate a meeting summary from a meeting transcript using OpenAI's ChapGPT API.
+
+## Usage
+
+Install the gem.
+
+`gem install summarize-meeting`
+
+### Commands
+
+Input a meeting transcript and output a summary.
+
+#### Usage
+
+`summarize-meeting transcript.txt`
+
+#### Options
+
+`-o, --output-file FILE` - Path to the output file. By default, it will be the `${input-file}-summary.txt`  
+`-k, --openai-key KEY` - The OpenAI API Key. Can also be set via `ENV["OPENAI_KEY"]`.  
+`-g, --openai-org ORG` - The OpenAI org id. It can also be set via `ENV["OPENAI_ORG"]` .  
+
+## Example
+
+This is an example summary from a 3,700 word transcript from a 30 minute meeting.
+
+> Summary of the Meeting:
+> 
+> The meeting started with discussions focusing on individual progress reports, where Sean shared updates on profit-related work, consolidating estimates, and improving Hey Kayla by creating a new JSON structure for Chat GPT messages API. Colenso reported finding a bug in the JavaScript Google Maps API and not making progress on push notifications. Meanwhile, Pankaj worked on bulk assign form handling, retaining time zone, and Benjamin faced some issues regarding database connection but eventually resolved the casting and integration issues.
+> 
+> In the open discussion session, the team discussed the confusion over parameters and importing across different branches. Benjamin shared about an open discussion necessary for a feature related to job production plans at ACME. Shirish discussed fixing the cycle time summary query, which used to take up to 6-7 seconds and now takes less than 1 second. Anish added a business unit filter feature to the time card and will work on page titles.
+> 
+> The meeting concluded with Sean hoping to have success with profit stuff before Monday's call.
+> 
+> Attendees:
+> 
+> - Sean
+> - Colenso
+> - Pankaj
+> - Benjamin
+> - Shirish
+> - Anish
+> 
+> Action Items:
+> 
+> - Sean to continue working on profit-related things before the next call.
+> - Colenso to continue working on resolving the bug in the Google Maps API and push notifications.
+> - Pankaj to continue working on bulk assign form handling and to keep the time zone option.
+> - Benjamin to finalize the release notes for the open discussion at  and to continue working on resolving the database connection issues.
+> - Shirish to further improve the query for cycle time summary and reduce the processing time.
+> - Anish to improve page titles and add filters for the time card. 
+> 
+> Detailed notes:
+> 
+> Sean:
+> 
+> - Reported progress on Prophet-related things
+> - Worked on consolidating all effective estimates
+> - Worked on improving Hey Kayla by creating a new JSON structure for Chat GPT messages API
+> 
+> Colenso:
+> 
+> - Found a bug in the JavaScript Google Maps API
+> - Has not made progress on push notifications yet
+> 
+> Pankaj:
+> 
+> - Worked on bulk assign form handling
+> - Allowed the option to retain the time zone
+> 
+> Benjamin:
+> 
+> - Faced difficulty regarding database connection issues in Heroku
+> - Resolved casting issues by switching to a different cash store
+> - Listed an interesting issue regarding split loads with more than one material transaction during the same trip
+> - Mentioned an integration issue with ACME
+> - Shared that there is an open discussion at ACME and he needs assistance with the release notes for a feature related to job production plans
+> 
+> Open Discussion:
+> 
+> - The team discussed the confusion over parameters and importing across different branches.
+> 
+> Shirish:
+> 
+> - Discussed fixing the cycle time summary query, which used to take up to 6-7 seconds and now takes less than 1 second
+> 
+> Anish:
+> 
+> - Added a business unit filter feature to the time card
+> - Will work on improving page titles
+
+## Credits
+
+This gem was created by XBE.

data/bin/summarize-meeting

@@ -1,8 +1,7 @@
 #!/usr/bin/env ruby
 
 require "optparse"
-
-require_relative "../lib/meeting"
+require "summarize-meeting"
 
 def main
   options = {}
@@ -19,8 +18,8 @@ def main
       options[:output_file] = file
     end
 
-    if ENV["OPENAI_KEY"]
-      options[:openai_key] = ENV["OPENAI_KEY"]
+    if ENV["OPENAI_API_KEY"]
+      options[:openai_key] = ENV["OPENAI_API_KEY"]
     end
 
     opts.on("-k", "--openai-key KEY", "The OpenAI API key to use") do |key|
@@ -36,8 +35,8 @@ def main
     end
   end.parse!
 
-  Ai.access_token = options[:openai_key] if options[:openai_key]
-  Ai.organization_id = options[:openai_org] if options[:openai_org]
+  SummarizeMeeting::Ai.access_token = options[:openai_key] if options[:openai_key]
+  SummarizeMeeting::Ai.organization_id = options[:openai_org] if options[:openai_org]
 
   if ARGV.length != 1
     puts "Error: You must specify a transcript file to summarize."
@@ -47,7 +46,7 @@ def main
   transcript_file = ARGV[0]
   transcript = File.read(transcript_file)
 
-  meeting = Meeting.new(transcript)
+  meeting = SummarizeMeeting::Meeting.new(transcript)
   summary = meeting.summarize
   summary_file_name = if options[:output_file]
     options[:output_file]
@@ -59,6 +58,6 @@ def main
   File.write(summary_file_name, summary)
 end
 
-if __FILE__ == $0
+if __FILE__.to_s.end_with?("summarize-meeting") && $0.to_s.end_with?("summarize-meeting")
   main
 end

data/lib/summarize-meeting/ai.rb

@@ -0,0 +1,28 @@
+require "openai"
+
+module SummarizeMeeting
+  module Ai
+    @@access_token = ENV["OPENAI_KEY"]
+    @@organization_id = ENV["OPENAI_ORG"]
+
+    def self.client
+      OpenAI::Client.new(access_token: access_token, organization_id: organization_id)
+    end
+
+    def self.access_token
+      @@access_token
+    end
+
+    def self.organization_id
+      @@organization_id
+    end
+
+    def self.access_token=(token)
+      @@access_token = token
+    end
+
+    def self.organization_id=(id)
+      @@organization_id = id
+    end
+  end
+end

data/lib/summarize-meeting/meeting.rb

@@ -0,0 +1,130 @@
+require "json"
+require "mustache"
+require "openai"
+
+module SummarizeMeeting
+  class Meeting
+    LINE_SUMMARY_PROMPT_TEMPLATE = [
+      {
+        role: "system",
+        content: "You are an assistant summarizing a meeting.",
+      },
+      {
+        role: "system",
+        content: "The transcript of the meeting is split into {{chunkCount}} chunks. This is the {{chunkIndex}} chunk.",
+      },
+      {
+        role: "assistant",
+        content: "Please provide me with the next chunk of the transcript.",
+      },
+      {
+        role: "user",
+        content: "{{chunk}}",
+      }
+    ]
+
+    CONSOLIDATED_SUMMARY_PROMPT_TEMPLATE = [
+      {
+        role: "system",
+        content: "You are an assistant summarizing a meeting.",
+      },
+      {
+        role: "system",
+        content: "Notes about the meeting have been compiled.",
+      },
+      {
+        role: "system",
+        content: <<~CONTENT
+          Your job is to write a thorough summary of the meeting.
+          The summary should start with a brief overview of the meeting.
+          The summary should be detailed and should extract any action items that were discussed.
+          The summary should be organized into sections with headings and bullet points.
+          The summary should include a list of attendees.
+          The order of the sections should be overview, attendees, action items, and detailed notes by topic.
+        CONTENT
+      },
+      {
+        role: "assistant",
+        content: "Please provide me with notes from the meeting.",
+      },
+      {
+        role: "user",
+        content: "{{notes}}",
+      }
+    ]
+
+    def initialize(transcript)
+      @transcript = transcript
+    end
+
+    attr_reader :transcript
+
+    def summarize
+
+      # Step 1. Split the transcript into lines.
+      lines = transcript.split("\n")
+
+      # Step 2. Calculate the maximum chunk size in words.
+      max_total_tokens = 4000
+      response_token_reserve = 500
+      template_tokens = LINE_SUMMARY_PROMPT_TEMPLATE.map { |line| line[:content].split.size }.sum
+      max_chunk_tokens = max_total_tokens - response_token_reserve - template_tokens
+      words_per_token = 0.7
+      max_chunk_word_count = max_chunk_tokens * words_per_token
+
+      # Step 3. Split the transcript into equally sized chunks.
+      chunks = split_lines_into_equal_size_chunks(lines, max_chunk_word_count)
+
+      # Step 4. Summarize each chunk.
+      previous_chunks_summary = ""
+      chunks.each_with_index do |chunk, chunk_index|
+        chunk_summary = summarize_chunk(chunk, chunk_index, chunks.size, previous_chunks_summary)
+        previous_chunks_summary += chunk_summary
+      end
+
+      # Step 5. Write a consolidated summary.
+      consolidated_template = CONSOLIDATED_SUMMARY_PROMPT_TEMPLATE
+      prompt = Mustache.render(consolidated_template.to_json, { notes: previous_chunks_summary.to_json })
+      messages = JSON.parse(prompt)
+      response = SummarizeMeeting::Ai.client.chat(
+        parameters: {
+          model: "gpt-3.5-turbo",
+          messages: messages,
+        }
+      )
+      response.dig("choices", 0, "message", "content")
+    end
+
+    def summarize_chunk(chunk, chunk_index, chunk_count, previous_chunks_summary)
+      template = LINE_SUMMARY_PROMPT_TEMPLATE
+      prompt = Mustache.render(template.to_json, { chunkCount: chunk_count, chunkIndex: chunk_index + 1, chunk: chunk.join("\n").to_json })
+      messages = JSON.parse(prompt)
+
+      response = SummarizeMeeting::Ai.client.chat(
+        parameters: {
+          model: "gpt-3.5-turbo",
+          messages: messages,
+        }
+      )
+      response.dig("choices", 0, "message", "content")
+    end
+
+    def split_lines_into_equal_size_chunks(lines, max_chunk_word_count)
+      chunks = []
+      chunk = []
+      chunk_word_count = 0
+      lines.each do |line|
+        line_word_count = line.split.size
+        if chunk_word_count + line_word_count > max_chunk_word_count
+          chunks << chunk
+          chunk = []
+          chunk_word_count = 0
+        end
+        chunk << line
+        chunk_word_count += line_word_count
+      end
+      chunks << chunk
+      chunks
+    end
+  end
+end

data/lib/summarize-meeting/version.rb

@@ -0,0 +1,3 @@
+module SummarizeMeeting
+  VERSION = "0.2.0"
+end

data/lib/summarize-meeting.rb

@@ -0,0 +1,6 @@
+require_relative "./summarize-meeting/version"
+require_relative "./summarize-meeting/ai"
+require_relative "./summarize-meeting/meeting"
+
+module SummarizeMeeting
+end

metadata

@@ -1,7 +1,7 @@
 --- !ruby/object:Gem::Specification
 name: summarize-meeting
 version: !ruby/object:Gem::Version
-  version: 0.1.1
+  version: 0.2.0
 platform: ruby
 authors:
 - Sean Devine
@@ -25,7 +25,7 @@ dependencies:
       - !ruby/object:Gem::Version
         version: '0'
 - !ruby/object:Gem::Dependency
-  name: dotenv
+  name: ruby-openai
   requirement: !ruby/object:Gem::Requirement
     requirements:
     - - ">="
@@ -39,7 +39,7 @@ dependencies:
       - !ruby/object:Gem::Version
         version: '0'
 - !ruby/object:Gem::Dependency
-  name: ruby-openai
+  name: mustache
   requirement: !ruby/object:Gem::Requirement
     requirements:
     - - ">="
@@ -53,13 +53,13 @@ dependencies:
       - !ruby/object:Gem::Version
         version: '0'
 - !ruby/object:Gem::Dependency
-  name: mustache
+  name: rspec
   requirement: !ruby/object:Gem::Requirement
     requirements:
     - - ">="
       - !ruby/object:Gem::Version
         version: '0'
-  type: :runtime
+  type: :development
   prerelease: false
   version_requirements: !ruby/object:Gem::Requirement
     requirements:
@@ -67,7 +67,7 @@ dependencies:
       - !ruby/object:Gem::Version
         version: '0'
 - !ruby/object:Gem::Dependency
-  name: rspec
+  name: guard-rspec
   requirement: !ruby/object:Gem::Requirement
     requirements:
     - - ">="
@@ -81,7 +81,7 @@ dependencies:
       - !ruby/object:Gem::Version
         version: '0'
 - !ruby/object:Gem::Dependency
-  name: guard-rspec
+  name: vcr
   requirement: !ruby/object:Gem::Requirement
     requirements:
     - - ">="
@@ -95,7 +95,7 @@ dependencies:
       - !ruby/object:Gem::Version
         version: '0'
 - !ruby/object:Gem::Dependency
-  name: vcr
+  name: webmock
   requirement: !ruby/object:Gem::Requirement
     requirements:
     - - ">="
@@ -109,7 +109,7 @@ dependencies:
       - !ruby/object:Gem::Version
         version: '0'
 - !ruby/object:Gem::Dependency
-  name: webmock
+  name: dotenv
   requirement: !ruby/object:Gem::Requirement
     requirements:
     - - ">="
@@ -130,16 +130,21 @@ executables:
 extensions: []
 extra_rdoc_files: []
 files:
+- LICENSE.txt
+- README.md
 - bin/summarize-meeting
-- lib/ai.rb
-- lib/meeting.rb
+- lib/summarize-meeting.rb
+- lib/summarize-meeting/ai.rb
+- lib/summarize-meeting/meeting.rb
+- lib/summarize-meeting/version.rb
 homepage: 
-licenses: []
+licenses:
+- MIT
 metadata: {}
 post_install_message: 
 rdoc_options: []
 require_paths:
-- "."
+- lib
 required_ruby_version: !ruby/object:Gem::Requirement
   requirements:
   - - ">="

data/lib/ai.rb

@@ -1,26 +0,0 @@
-require "openai"
-
-module Ai
-  @@access_token = ENV["OPENAI_KEY"]
-  @@organization_id = ENV["OPENAI_ORG"]
-
-  def self.client
-    OpenAI::Client.new(access_token: access_token, organization_id: organization_id)
-  end
-
-  def self.access_token
-    @@access_token
-  end
-
-  def self.organization_id
-    @@organization_id
-  end
-
-  def self.access_token=(token)
-    @@access_token = token
-  end
-
-  def self.organization_id=(id)
-    @@organization_id = id
-  end
-end

data/lib/meeting.rb

@@ -1,130 +0,0 @@
-require "json"
-require "mustache"
-require "openai"
-
-require_relative "./ai"
-
-class Meeting
-  LINE_SUMMARY_PROMPT_TEMPLATE = [
-    {
-      role: "system",
-      content: "You are an assistant summarizing a meeting.",
-    },
-    {
-      role: "system",
-      content: "The transcript of the meeting is split into {{chunkCount}} chunks. This is the {{chunkIndex}} chunk.",
-    },
-    {
-      role: "assistant",
-      content: "Please provide me with the next chunk of the transcript.",
-    },
-    {
-      role: "user",
-      content: "{{chunk}}",
-    }
-  ]
-
-  CONSOLIDATED_SUMMARY_PROMPT_TEMPLATE = [
-    {
-      role: "system",
-      content: "You are an assistant summarizing a meeting.",
-    },
-    {
-      role: "system",
-      content: "Notes about the meeting have been compiled.",
-    },
-    {
-      role: "system",
-      content: <<~CONTENT
-        Your job is to write a thorough summary of the meeting.
-        The summary should start with a brief overview of the meeting.
-        The summary should be detailed and should extract any action items that were discussed.
-        The summary should be organized into sections with headings and bullet points.
-        The summary should include a list of attendees.
-        The order of the sections should be overview, attendees, action items, and detailed notes by topic.
-      CONTENT
-    },
-    {
-      role: "assistant",
-      content: "Please provide me with notes from the meeting.",
-    },
-    {
-      role: "user",
-      content: "{{notes}}",
-    }
-  ]
-
-  def initialize(transcript)
-    @transcript = transcript
-  end
-
-  attr_reader :transcript
-
-  def summarize
-
-    # Step 1. Split the transcript into lines.
-    lines = transcript.split("\n")
-
-    # Step 2. Calculate the maximum chunk size in words.
-    max_total_tokens = 4000
-    response_token_reserve = 500
-    template_tokens = LINE_SUMMARY_PROMPT_TEMPLATE.map { |line| line[:content].split.size }.sum
-    max_chunk_tokens = max_total_tokens - response_token_reserve - template_tokens
-    words_per_token = 0.7
-    max_chunk_word_count = max_chunk_tokens * words_per_token
-
-    # Step 3. Split the transcript into equally sized chunks.
-    chunks = split_lines_into_equal_size_chunks(lines, max_chunk_word_count)
-
-    # Step 4. Summarize each chunk.
-    previous_chunks_summary = ""
-    chunks.each_with_index do |chunk, chunk_index|
-      chunk_summary = summarize_chunk(chunk, chunk_index, chunks.size, previous_chunks_summary)
-      previous_chunks_summary += chunk_summary
-    end
-
-    # Step 5. Write a consolidated summary.
-    consolidated_template = CONSOLIDATED_SUMMARY_PROMPT_TEMPLATE
-    prompt = Mustache.render(consolidated_template.to_json, { notes: previous_chunks_summary.to_json })
-    messages = JSON.parse(prompt)
-    response = Ai.client.chat(
-      parameters: {
-        model: "gpt-3.5-turbo",
-        messages: messages,
-      }
-    )
-    response.dig("choices", 0, "message", "content")
-  end
-
-  def summarize_chunk(chunk, chunk_index, chunk_count, previous_chunks_summary)
-    template = LINE_SUMMARY_PROMPT_TEMPLATE
-    prompt = Mustache.render(template.to_json, { chunkCount: chunk_count, chunkIndex: chunk_index + 1, chunk: chunk.join("\n").to_json })
-    messages = JSON.parse(prompt)
-
-    response = Ai.client.chat(
-      parameters: {
-        model: "gpt-3.5-turbo",
-        messages: messages,
-      }
-    )
-    response.dig("choices", 0, "message", "content")
-  end
-
-  def split_lines_into_equal_size_chunks(lines, max_chunk_word_count)
-    chunks = []
-    chunk = []
-    chunk_word_count = 0
-    lines.each do |line|
-      line_word_count = line.split.size
-      if chunk_word_count + line_word_count > max_chunk_word_count
-        chunks << chunk
-        chunk = []
-        chunk_word_count = 0
-      end
-      chunk << line
-      chunk_word_count += line_word_count
-    end
-    chunks << chunk
-    chunks
-  end
-end