This article is for Content Leaders and Strategists who drive tech doc strategies for their organisations. It discusses some uncomfortable questions that AI leaders must answer while implementing AI in their documentation systems. Let’s go!

Ask the Whys Before You Start

Why do you need AI-driven tech doc strategies? Because:

• AI is the buzzword.
• Everywhere there is AI.
• AI is the present and the future.
• Leadership has asked you to use AI.
  

If you are experimenting with AI for the above reasons, my dear friend, dive deeper before implementing AI in your tech doc processes. You will do more harm than good—for your docs and for your organisation. Take a pause. Do you really need AI?

A lot needs to be reconsidered before diving into AI-based strategies.

If you want to use AI because:

• AI can help reduce redundant work.
• AI can analyse large data sets and draw conclusions that help writers make better decisions.
• AI can assist where tech writers have dependencies on other teams.
• AI empowers tech writers to do more in less time.
  

You must explore further. AI can perhaps help.

How to Build AI-Based Content Strategies

You build AI-based content strategies the same way you build regular ones—but yes, there are some differences, which we’ll discuss further in this blog.

Think Like a Product Manager

In my opinion, a Content Strategist or Content Leader is the Product Manager for Tech Docs. They define the organisation’s style guides, processes, tools, and documentation systems. They work with cross-functional teams to design the content architecture, format, layout, functionalities, and design of docs.

So, if you want to implement AI, you need to clearly define the whys and then the whats before you start. Decide on the data metrics that will be used to evaluate the success or failure of your doc strategies.

Identify What Problems You Are Trying to Solve

Now that you are convinced AI can help, identify where it can help. Examples include:

• Creating the first draft
• Reviewing and polishing content
• Creating FAQs
• Creating diagrams
• Helping with research

And much more… to begin with, you can do the following:

• Identify where your team spends most of its time and how much of it can be automated. For example, at Razorpay, we found that our team spends ~50% of its time on content creation. We tried automating that with AI.
• Understand the quality of the automated content. Do they need to spend more time reviewing it?
• What additional things can your tech writers do with AI? For example, at Razorpay, we use Cursor rules to make code changes, which previously required developer support.
  

Look at Market Leaders and Do Market Research

Check how others are adopting AI in their documentation processes. Explore their documentation systems to see AI in action—these can often be good indicators.

But remember: don’t do things just because others are doing them. Do them only if you see value. If their solutions are failing, learn from them.

Define Success Criteria

How do you measure the success of AI implementation? For example:

• Improved efficiency: Earlier, a large task took 7 days; now, with AI, it takes half a day.
• Improved content quality: Embedding AI in your CMS can catch spelling or grammar errors instantly. If earlier, ‘x’ language errors were caught in editorial reviews, now with AI, there are none.
• Ability to do more: Tech writers once struggled with video creation because traditional tools were time-consuming. Now, AI creates videos—writers just provide the screen recordings and scripts.
  

And there can be many more, depending on your use case.

Secure Budgets. Get Compliance Approval.

Get budget approval from leadership before you start experimenting with AI.

• Research tool requirements and pricing.
• Stack rank your asks—not everything will get approved.
• Look at ROI. AI tools, especially growth plans for teams, are expensive. Do the maths: how much are you saving versus how much you’ll spend?
  

Understand compliance requirements before going deeper into AI experiments.

I have learnt this the hard way. As a tech-savvy and an over-enthusiast, I often beat myself up for spending hours experimenting with AI tools before securing compliance approval.

AI tools may look magical and an excellent fit. But ask:

• What data are they collecting?
• Are they using your data to train their LLMs?
• Do they have the required compliance certifications?
  

Check these before investing much time. Many tools do not provide sufficient data security unless you purchase enterprise plans, which are very expensive.

Understand AI. Take a Hands-on Approach. Build a Team.

AI is evolving fast. Every other day, a new LLM, model, or product launch occurs. It’s overwhelming.

• Enrol in basic AI courses to understand key concepts and terminology.
• Attend webinars and community sessions to see how the industry is adopting AI.
• Read articles and listen to podcasts to stay updated.
  

I always recommend a hands-on approach. It gives you confidence and control. Install a few free versions or starter subscriptions—the learning is priceless.

Of course, you cannot try everything yourself. That’s why I recommend building a team to evaluate and find the best fits for your use case. Split the work, track progress, and drive execution together.

Final Words

Creating AI-driven tech doc strategies is exciting, overwhelming, rewarding, and sometimes frustrating. If you are well-prepared, your team and organisation’s chances of success are higher. Use AI to simplify tasks, reduce redundancy, explore areas previously out of reach, and do more!