Skip to main content
Apragya AIFree quotation builderQuotation builderQuote builderno signup
How-to guide

How to Automate Knowledge Base Articles in an ERP: Answers Your Support Desk Can Find and Trust

For support heads and service managers: write each answer once, keep it current, and put it in front of agents and customers at the moment they need it.

A support head placing a binder on a neat bookshelf in a service centre

How to automate knowledge base articles is a question support teams ask when the same five answers are typed forty times a week. The knowledge exists, but it lives in a senior agent's head, an old Word file and a WhatsApp group.

New joiners interrupt the senior, the senior answers from memory, and customers sometimes get terms that changed last quarter. Customers who would happily find the answer themselves raise a ticket because there is nothing to search.

This guide covers the steps to a working knowledge base, an illustrative example at a Mysuru inverter and battery dealer, and how Apragya's support app handles articles.

How to automate knowledge base articles, step by step

A knowledge base is a library of approved answers for agents and customers. Automating it is less about writing articles by machine and more about making them easy to create, easy to find and easy to keep current.

The best source is your resolved tickets. When an agent writes a clear answer, turn it into a draft article, give it a category and tags, and publish it once checked. Decide whether customers may see it or whether it stays internal.

Then put the articles where work happens. Suggest them on new tickets, offer them in a website help widget, and track views and helpful votes so weak articles are rewritten and outdated ones archived.

  1. Pick the questions your desk answers most often.
  2. Draft an article from the best ticket reply for each.
  3. Assign categories and tags, then publish.
  4. Mark each article as public or internal.
  5. Turn on article suggestions for agents and a help widget for customers.
  6. Review usage monthly and archive what is out of date.

What an unwritten knowledge base costs

Seniors repeating answers

Experienced staff spend the day answering new joiners.

Outdated terms

An old warranty file is forwarded after the policy changed.

Avoidable tickets

Customers raise tickets for questions an article could answer.

Nothing to search

Answers sit in chats that nobody else can find.

An example: warranty questions at a Mysuru battery dealer

Take an illustrative inverter and battery dealer in Mysuru with a desk of three and eight field technicians. Warranty and water top-up questions came in daily, and each person explained them differently.

From one resolved ticket to a reused answer

  1. Ticket resolved

    A clear warranty answer written by the desk.

  2. Article drafted

    Turned into an article in a warranty category.

  3. Published

    Status moved from draft to published.

  4. Suggested on tickets

    Offered to agents on similar tickets.

  5. Found by customers

    Searched from the website help widget.

The article earns its keep when it is suggested on the next similar ticket.

Within a month the desk had twenty-five published articles. Agents picked suggested articles instead of typing from memory, technicians searched them from their phones before calling the office, and customers who asked in Kannada still found the English battery care article.

A technician reading a card from a box beside a battery on a counter
The battery care article, suggested on the ticket and read by the technician before the site visit.

How Apragya runs the help desk knowledge base

Step 1 of 5

Organise by category

Articles sit in categories with tags, so the library stays browsable.

  1. Organise by category: Articles sit in categories with tags, so the library stays browsable.
  2. Draft, publish, archive: Each article moves from draft to published, and old ones are archived.
  3. Public or internal: Mark an article for customers or keep it for the team.
  4. Suggested on tickets: AI suggests relevant articles while an agent works a ticket.
  5. Search in any language: Semantic search runs alongside keyword search, across languages.

Self service knowledge base: shared drive or built into support

A shared drive stores files but cannot say which one is current, and it is invisible when an agent is working a ticket. Articles that sit beside the tickets they answer, with views and helpful votes counted, show which answers work and which need rewriting.

Knowledge approaches, compared with Apragya

FeatureWord files and a WhatsApp groupApragya knowledge base
Current versionSeveral undated copiesOne article with a status
Finding an answerAsk a seniorSearch the library
While working a ticketSwitch windows and huntArticles suggested on the ticket
CustomersCannot see itPublic articles in the help widget
UsefulnessUnknownViews and helpful votes counted

Checked against public information on 16 September 2026.

What sets Apragya apart on support article suggestions

Across languages

A question in one language can find an article written in another.

Exact matches kept

Keyword search still finds part numbers and error codes.

Help widget

Public articles can be searched from a widget on your website.

Bulk import

Existing articles can be imported from a CSV file.

Usage counted

Views and helpful votes show which articles work.

Phone-friendly

Technicians search articles in the browser on their phones.

Articles pair with short answers in canned replies and are used when working tickets. See the knowledge base articles feature and best customer support software.

Getting started

  1. List the ten questions your desk answers most often.
  2. Write one article for each, starting from the best ticket reply.
  3. Group them into a few categories with tags.
  4. Mark which articles customers may see.
  5. Check views and helpful votes after a month and rewrite the weak ones.

Twenty good articles cover most of what a new joiner asks in the first month.

What it costs to start

INR 3,999
Starter, a month
Up to 10 users, CRM and HR plus Support and Ticketing
2,000
AI Credits a month
Included on Starter
Free
Up to 3 users
CRM, HR and Vendor Onboarding

Automating knowledge base articles, answered

Can customers search the articles?
Yes. Articles marked public can be searched from the website help widget; internal ones stay with the team.
Does search work if customers use their own language?
Semantic search matches meaning across languages, and keyword search still runs for exact terms.
Can we import articles we already have?
Yes. Articles can be imported in bulk from a CSV file.
How do agents find the right article on a ticket?
AI can suggest relevant articles for the ticket being worked, and the agent decides what to send.
What does it cost to start?
Support and Ticketing is available from the Starter plan, INR 3,999 a month for up to 10 users, which includes CRM and HR plus a choice of enterprise apps and 2,000 AI Credits a month. Pro, at INR 7,999 a month, covers 25 users and every enterprise app.

Write the answer once

Publish your most common answers and let tickets and customers find them.