Why Your AI Results Are Completely Off

Welcome to Issue #1 of the ElevateX Newsletter. Each week, one practical AI skill to help you get ahead. Takes 4 minutes to read.

Let me guess. You've used ChatGPT to write an email, summarize notes, or maybe even fix your resume. And sometimes it gives brilliant answers. Other times? Not even close.

Here's the thing - the difference between a helpful AI response and a completely off one has nothing to do with which AI tool you're using. It has everything to do with how you talk to it.

That skill of talking to AI properly? It has a name. It's called Prompt Engineering.

And if you're a fresher in India right now, this might be the most practical skill you pick up this year.

Forget the fancy name for a second. Prompt engineering is simply the skill of giving clear, specific instructions to an AI tool so it gives you exactly what you need.

Think of it like this. You walk into a restaurant and say "give me something nice." The waiter is confused. You'll probably get something random. But if you say "I want a spicy paneer dish, not too oily, with butter naan on the side" - you get exactly what you wanted.

That's prompt engineering. Being specific with AI instead of being vague.

Here's a real example:

Bad prompt: "Write me a cover letter."

Good prompt: "Write a cover letter for a fresher applying to a marketing analyst role at Swiggy. I have a BBA from Pune University, did a 3-month internship in social media marketing, and I'm good with Excel and Google Analytics. Keep it under 200 words and make it sound confident but not over the top."

Same AI tool. Completely different output. The only thing that changed was the instruction.

There's actually a simple formula behind that good prompt. Most strong prompts follow this structure:

Role: Who should the AI act as? ("Act as a senior HR manager...")

Task: What do you need? ("...review my resume and point out weak areas")

Context: What's the background? ("...I'm a fresher from a tier 2 college applying to IT companies in Pune")

Format: How do you want the output? ("...give me a bullet list with specific suggestions, not generic advice")

Put all four together and you go from getting a generic response to getting something you can actually use. Save this formula. You'll use it every week.

Three reasons this matters for freshers right now.

First, Indian companies are already hiring for this. As of early 2025, Naukri.com lists over 12,000 job postings that mention prompt engineering as a skill. And these aren't just deep-tech roles. Companies like Accenture, Deloitte, and even startups like TensorGo and upGrowth are looking for people who can work with AI tools across marketing, content, operations, and more. Even Internshala has started listing prompt engineering roles specifically for freshers. This is not a future thing. It's happening right now.

Second, it makes you faster at everything. Writing emails, preparing for interviews, summarizing research, building presentations - all of it gets 3 times faster when you know how to prompt properly. Your batchmates are spending 2 hours on what you could finish in 30 minutes.

Third, it's free to learn. You don't need a course, a certificate, or a high-end laptop. You just need access to any free AI tool (ChatGPT, Gemini, Claude - all have free versions) and the willingness to practice.

Open ChatGPT or any AI tool and try this prompt:

"I'm a final year [your degree] student in [your city]. I want to apply for [type of role] jobs. Give me 5 specific things I should learn or do in the next 30 days to become a stronger candidate. Be practical and India-specific."

See the difference between this and just typing "how to get a job"? That's effective prompt engineering in action.

Prompt engineering is not about memorizing magic sentences. It's about thinking clearly about what you want before you ask. The better your input, the better your output. That's true for AI, and honestly, that's true for life too.

Next issue, we're covering the 5 AI skills that Indian recruiters are actually looking for right now. Not the theoretical stuff - the ones that actually show up in job descriptions and interviews.

If this was useful, forward it to that one friend who still copies ChatGPT outputs without editing them. They need this.

Got feedback for me? Want me to review one of your prompts and make it more effective? Just reply or write to me at [email protected] - I read every message personally.

Until next week,

Vicky