Key Takeaways from the Prompt Engineering Session with Charu Misra-Batuhan
Most professionals believe they’re using AI effectively. They open a chatbot, type a quick instruction, and accept the first response. However, Charu Misra-Batuhan, AI enthusiast and Digital Marketing Consultant of John Clements explained that this habit is exactly what limits productivity. The truth is simple: the quality of your output depends entirely on the quality of your input.
The good news? It’s fixable — and it starts with structured prompt engineering techniques.
The CRAFT Formula: A Better Way to Talk to AI
Charu introduced the CRAFT formula, a five-part framework that transforms vague requests into precise, high-quality prompts. Each letter represents a critical element that professionals often overlook.
| ELEMENT | WHAT IT MEANS |
| C — Context | Set the scene. Who are you? What’s the situation? |
| R — Role | Assign a persona. Tell AI to ‘Act as a Recruiter’ or ‘Act as a Financial Analyst.’ |
| A — Action | State the specific task clearly and directly. |
| F — Format | Define the output — bullet points, table, email, a specific length. |
| T — Tone | Specify the voice — professional, friendly, concise, or empathetic. |
To demonstrate, she compared two prompts:
- “Write a job offer email.” → Generic, uninspired output.
- “As a senior recruiter, write a job offer email for a marketing role including compensation, benefits, and acceptance deadline.” → Personalized, ready-to-send deliverable.
This single shift in approach shows how prompt engineering techniques elevate AI from novelty to genuine productivity tool.
How AI Actually Processes Your Words
Understanding the mechanics of AI clarifies why precision matters. Charu outlined four stages:
- Tokenization – Input is broken into tokens (words or parts of words).
- Pattern Matching – AI scans for patterns based on training data.
- Generation – It predicts the most useful next word step by step.
- Output – The final response is assembled and delivered.
The key insight: AI has no intuition. It cannot infer missing context or “read between the lines.” Consequently, every detail you omit forces AI to guess — often incorrectly.
Three Mistakes That Quietly Cost You Results
Charu highlighted three common errors that undermine AI responses:
- Being Too Vague
- Example: “Rephrase this paragraph.”
- Fix: Specify tone, format, audience, and role.
- Not Iterating
- Many users accept the first draft.
- Fix: Treat prompting as a conversation. Refine outputs step by step.
- Skipping the Role
- Example: “Create a report.” → Generic.
- Fix: “Act as a hiring manager and create a board-ready report on candidate pipeline gaps.”
These corrections align with effective prompt engineering techniques that consistently improve results.
The Smart Intern Mindset
Charu’s analogy resonated strongly: think of AI as a smart intern. It’s capable and eager, but only if briefed properly. You wouldn’t hand a new team member a half-formed request and expect perfection. In contrast, clear instructions transform AI into a reliable productivity partner.
That means:
- Be specific.
- Don’t settle for the first answer.
- Double-check outputs before use.
Every great result starts with a great prompt. This mindset shift is the foundation of modern prompt engineering techniques.
Want More Insights Like This?
This session is one of many expert-led conversations hosted by John Clements to help professionals thrive in a rapidly evolving workplace. Whether you’re sharpening AI skills, navigating talent challenges, or staying ahead of industry trends, we deliver the insights that matter.
Explore upcoming events, resources, and opportunities at John Clements.
External References for Authority
- https://www.nist.gov (National Institute of Standards and Technology)
- https://www.gov.ph (Official Philippine Government Portal)
