The Fifth Elephant 2023 Winter

The Fifth Elephant 2023 Winter

On the engineering and business implications of AI & ML



This video is for members only

Sumit Saha


Automating Knowledge Extraction for Training Content Generation with LLMs @ SquadStack

Submitted Nov 13, 2023


In traditional BPOs, training sales representatives is challenging, primarily due to limited visibility into the universe of knowledge that the representatives have to stay on top of. This limitation often results in suboptimal sales interactions with customers. Furthermore, the creation of thorough and timely training materials is a time-consuming endeavor, even before the actual training can commence. Consequently, this extended preparatory phase delays sales representatives from promptly engaging in customer calls on behalf of the business, resulting in missed opportunities.

As a solution to this problem, SquadStack is in the process of creating an in-house tool, we call Intel Assist, to streamline the extraction of content, for faster and accurate generation and querying of training material for a variety of businesses at a very low cost. This as a result, is empowering our sales representatives to access and utilize this information seamlessly during customer calls.

At its core, we aim for source-agnostic knowledge integration, gathering information from a wide range of sources to enhance documentation and streamline the process of querying this knowledge. While we have initially applied this approach to enhance our internal business processes at SquadStack, it can also benefit other enterprises seeking to bolster their own knowledge management capabilities.


Step 1 - Knowledge Onboarding

  • Extracting asked Questions, Questions & Answers (QnAs), and Summaries from a variety of raw data sources, including Call Recordings, Word Documents, PDFs, Spreadsheets, and Presentations.
  • Our comprehensive logging of each question and objection during calls has resulted in a valuable data repository, opening the door to a multitude of analytical insights.
  • By identifying frequently asked questions and objections, we equip our representatives with the means to improve their preparedness and performance.

Step 2 - Knowledge Sufficiency

  • Automatically generate a report assessing the alignment of the extracted knowledge with periodic changes and new requirements.
  • We can precisely discern which of the current questions being posed can be addressed using our existing knowledge base and which ones cannot.
  • Subsequently, we can present these specific questions to our customers, gather their responses, and incorporate the newfound insights back into our knowledge repository.

Step 3 - Knowledge Search

  • Swiftly interrogate an extensive knowledge database spanning multiple businesses.
  • This proves highly effective when sales representatives require real-time access to information that transcends conventional keyword matching.



  • Reduction in time required to acquire knowledge
  • Reduction in the expenses to qcquire knowledge
  • Accelerated onboarding for new customers, leading to quicker Return on Investment (ROI) delivery for our clients.

Flow of Presentation

  • SquadStack’s Introduction (1-2 mins)

  • Product Introduction (5-8 mins)

    • Industrial Problems (3-5 mins)
    • Overview of our solution - Intel Assist (2-3 mins)
    • Introduce ASR as previously showcased in Fifth Elephant Monsoon ‘23 (2-3 mins)
  • Features (7-12 mins)

    • Knowledge Onboarding (3-5 mins)
    • Knowledge Sufficiency (3-5 mins)
    • Knowledge Search (1-2 mins)
  • Application’s Lifecycle (3-5 mins)

    • What is the direct output of Intel Assist? (Streamlit App)
    • How do we provide training?
    • Impact (1-2 mins)
    • Time Saved
    • How can people make use of a tool like this
  • Audience QnA


Sumit Saha


{{ gettext('Login to leave a comment') }}

{{ gettext('Post a comment…') }}
{{ gettext('New comment') }}
{{ formTitle }}

{{ errorMsg }}

{{ gettext('No comments posted yet') }}

Hybrid access (members only)

Hosted by

All about data science and machine learning

Supported by


Providing all founders, at any stage, with free resources to build a successful startup.