← Blog
GENERATIVE AI · DATA ANALYTICS · INDIA 2026

Generative AI for Data Analysts India 2026
ChatGPT, Copilot & AI Tools — Practical Guide with Real Prompts

Generative AI is changing how data analysts work — not by replacing them, but by handling the mechanical work so analysts can focus on judgment, context, and decisions. This guide shows exactly how to use AI tools in your daily analytics work, with real prompts you can copy and use today.

See Real Use Cases →AI + Analytics Training →

The honest picture: what AI changes and what it does not

✅ What AI does well for analysts
• Writing boilerplate SQL and explaining each clause
• Debugging formulas — Excel, DAX, Python
• Generating first-draft Python code from plain description
• Summarising analysis findings for non-technical audiences
• Explaining what an unfamiliar function does
• Suggesting chart types and dashboard layouts
• Reformatting data between structures
❌ What AI cannot replace
• Knowing which business question is worth asking
• Judging whether a finding is meaningful or noise
• Understanding data quality in your specific system
• Stakeholder relationships and trust
• Domain knowledge (industry, company, product)
• Catching when AI output is plausible but wrong
• Strategic recommendations with full business context

6 Real Use Cases — With Prompts You Can Copy Today

Every use case includes a real prompt template and an honest warning about where AI-generated output needs your verification.

🔍

AI-Assisted SQL Writing

WITHOUT AI
Write a complex multi-table JOIN with window functions from scratch — 20–40 minutes for an intermediate analyst.
WITH AI
Describe the query in plain English to ChatGPT. Get a working draft in 30 seconds. Review, adjust, and verify — 3–5 minutes total.
Prompt example:
"I have two tables: orders (order_id, customer_id, amount, order_date, city) and customers (customer_id, name, segment). Write a SQL query that:
1. Calculates total revenue per customer segment per month
2. Shows month-over-month growth using LAG()
3. Only includes the last 12 months of data
4. Orders by segment and month ascending
Use MySQL syntax."
⚠️ Verify: Always verify the output. AI-generated SQL can produce logically plausible but incorrect results, especially with NULL handling and date functions. Test on a sample dataset first.
📊

AI for DAX in Power BI

WITHOUT AI
Writing a complex DAX measure (rolling 3-month average, YoY comparison, or rank within category) requires deep CALCULATE and context knowledge.
WITH AI
Describe the business metric in plain language. Copilot or ChatGPT generates the DAX. You understand what it does, verify it on sample data, and deploy.
Prompt example:
"I am building a Power BI report for an Indian FMCG company. Write a DAX measure that:
- Calculates the 3-month rolling average of sales
- Uses Indian financial year (April–March)
- Filters to the selected region from a slicer called 'Region'
- Returns BLANK() if fewer than 3 months of data exist
Explain each function used."
⚠️ Verify: DAX is sensitive to filter context. Always test the measure across all slicer combinations — AI often misses edge cases around ALLSELECTED vs ALL.
🐍

AI for Python Data Cleaning

WITHOUT AI
Writing pandas code to handle messy Indian datasets (mixed date formats, ₹ symbols in amount columns, inconsistent city name spellings) is tedious.
WITH AI
Paste a sample of the messy data and describe the problem. Get working pandas code that handles the specific issues in your dataset.
Prompt example:
"I have a pandas DataFrame with these issues:
- 'amount' column has values like '₹1,50,000', '2.5L', '250000.00' — need them all as integers
- 'date' column mixes 'DD-MM-YYYY', 'DD/MM/YY', and 'Month DD YYYY' formats
- 'city' column has 'Delhi', 'delhi', 'New Delhi', 'DELHI' all meaning the same place
Write pandas code to clean all three columns. Include comments explaining each step."
⚠️ Verify: Test on the full dataset, not just the sample you showed the AI. Edge cases the AI did not see in the sample will break the code.
✍️

AI for Insight Narration

WITHOUT AI
Writing a clear, concise summary of analysis findings for a presentation or email to a non-technical manager takes significant time and writing skill.
WITH AI
Paste your key numbers and ask AI to write a 3-bullet executive summary or a 2-paragraph explanation tailored to the audience.
Prompt example:
"I have finished analysing our Q2 FY2026 sales data. Key findings:
- Revenue: ₹4.2Cr vs target ₹5Cr (84% attainment)
- North region: ₹1.8Cr, up 12% vs Q1
- South region: ₹0.9Cr, down 18% vs Q1 due to distributor transition
- Top SKU: Product X contributed 31% of total revenue
Write a 3-bullet executive summary for our VP Sales. Be direct, no jargon.
Then write a 1-paragraph recommendation for which region to prioritise in Q3."
⚠️ Verify: Always review AI-written narratives before sending. AI sometimes invents implications or adds hedging language that does not match your actual conclusions.
📗

AI for Excel Formula Help

WITHOUT AI
Debugging a nested XLOOKUP or a SUMIFS that returns wrong results requires systematic checking of each argument.
WITH AI
Paste the formula and describe what it should do. AI identifies the error and explains why, often in under a minute.
Prompt example:
"This Excel formula is returning 0 instead of the correct total:
=SUMIFS(Sales[Amount], Sales[City], A2, Sales[Date], ">="&DATE(2026,4,1), Sales[Date], "<="&DATE(2027,3,31))
My date column is formatted as DD-MM-YYYY text, not a real date. How do I fix this formula? Also explain why the current formula fails."
⚠️ Verify: AI is very good at formula debugging. However, it cannot see your actual data — always describe the data types and any formatting issues clearly in your prompt.
🤖

AI for Automated Report Commentary

WITHOUT AI
Monthly reports require the same narrative structure every month but with different numbers — writing the same commentary frame repeatedly is low-value work.
WITH AI
Build a prompt template with placeholders. Each month, fill in the numbers and run the prompt. AI generates the narrative in the same format, consistent and fast.
Prompt template (reuse every month):
"Monthly analytics report commentary. Audience: CFO and department heads.
Data for [MONTH] [YEAR]:
- Total orders: [X] vs [Y] last month ([Z]% change)
- Revenue: ₹[X] vs target ₹[Y] ([Z]% attainment)
- Top region: [X] with ₹[Y]
- Biggest drop: [Region/Category] down [Z]% due to [reason]
Write: (1) one-sentence headline, (2) 3-bullet performance summary, (3) one sentence on the biggest risk for next month. Keep it under 150 words total."
⚠️ Verify: Prompts for recurring reports should be version-controlled (save in a shared doc). This becomes institutional knowledge that any analyst on the team can use.

AI Tools for Data Analysts — Comparison 2026

Which tool to use for which task, with pricing in INR context.

ToolProviderCostBest For Data AnalystsIntegration
ChatGPT (GPT-4o)OpenAIFree tier + ₹1,600/mo ProSQL generation, Python debugging, insight narration, formula helpWeb, API, ChatGPT plugin ecosystem
Claude (claude.ai)AnthropicFree tier + paid ProLong document analysis, nuanced code explanation, structured outputsWeb, API, Claude for Teams
Microsoft CopilotMicrosoftIncluded in M365 Business plansExcel formula help, Power BI DAX, PowerPoint summaries — in-appNative in Excel, Power BI, Teams, Word
GitHub CopilotGitHub / OpenAI~₹850/moPython code completion, pandas, matplotlib, SQL in notebooksVS Code, JupyterLab, PyCharm
Gemini AdvancedGoogleIncluded in Google One AI plansGoogle Sheets formulas, BigQuery SQL, reading long reportsGoogle Workspace, Sheets, BigQuery
Perplexity AIPerplexityFree + Pro tierResearch with citations — understanding a dataset domain quicklyWeb, API

The 5 Rules of Good Prompts for Analytics Work

1
Give context first

Tell the AI what the data represents, what industry the company is in, and who will use the output. A prompt that says "I work in FMCG, analysing monthly distributor sales data" produces far better SQL than "I have sales data."

2
Be specific about the output format

Say "return as a SQL query with comments on each clause", "format as a 3-bullet summary", or "give the answer as a pandas one-liner". Vague requests produce vague answers.

3
Provide a sample of the data

Paste 5–10 rows of your actual data (remove sensitive info first). AI gives much more accurate code and formulas when it can see the real column names, data types, and values.

4
State constraints explicitly

If you need MySQL not PostgreSQL, Excel 2019 not Microsoft 365, or a formula without helper columns — say so. AI defaults to the most common version, which may not be what you need.

5
Always ask for an explanation

Add "explain what each part does" to every code prompt. This builds your own understanding and helps you catch errors. An analyst who understands the AI-generated code is far more valuable than one who just copies it.

Related AI & analytics guides
AI Tools for Data AnalystsGenAI Course in NoidaSQL Tutorial (for AI-assisted SQL)Python Tutorial

Frequently Asked Questions

Will Generative AI replace data analysts in India?

No — but it will replace data analysts who do not learn to use it. Generative AI automates the mechanical parts of analysis: writing boilerplate SQL, explaining code, generating first-draft Python scripts, and producing summary narratives. The parts it cannot replace are business judgment (knowing which question to ask), domain context (understanding why a metric moved), stakeholder communication (explaining the implication of a finding to a non-technical manager), and data quality judgment (knowing when an output is wrong despite looking correct). Analysts who use AI to do the mechanical work faster — and focus their own time on the judgment-heavy parts — become dramatically more productive and are exactly what Indian companies are trying to hire.

Which Generative AI tools should a data analyst in India learn in 2026?

The highest-value tools for data analysts in India in 2026 are: (1) ChatGPT or Claude — for generating and debugging SQL, explaining code, writing narrative summaries of analysis, and drafting presentation text; (2) Microsoft Copilot — integrated into Excel and Power BI, it can write DAX measures, suggest chart types, and summarise data from plain-language prompts; (3) GitHub Copilot — for Python code generation and debugging in VS Code; (4) Gemini Advanced — for long-document analysis, especially useful for reading research reports or vendor documentation quickly. Learning to write precise, context-rich prompts (prompt engineering) multiplies the value of all these tools.

What is prompt engineering and why do data analysts need it?

Prompt engineering is the skill of writing instructions to an AI model that consistently produce useful, accurate, and actionable outputs. For data analysts, good prompts include: the business context (what the data represents, what the company does), the specific task (write a SQL query to..., explain why this DAX formula returns..., summarise these findings for a non-technical stakeholder), and constraints (use only these columns, limit to 3 bullet points, format as a table). A vague prompt produces a generic answer. A well-structured prompt produces a working SQL query or a dashboard design you can actually use. Prompt engineering is now listed as a required or preferred skill in a growing number of data analyst job descriptions in India.

How is Generative AI changing data analyst job descriptions in India?

Indian data analyst job descriptions in 2026 increasingly mention: AI tools familiarity (ChatGPT, Copilot, Gemini), prompt engineering as a skill, experience with LLM-based reporting tools, and the ability to critically evaluate AI-generated outputs. At the same time, the bar for core analytical skills (SQL, Power BI, Python) has not dropped — companies expect analysts to use AI to work faster, not as a substitute for knowing the fundamentals. The emerging expectation is: strong foundational data skills plus the ability to leverage AI tools to multiply output — not one or the other.

EVIKA ACADEMY · NOIDA SECTOR 51 · AI + DATA ANALYTICS TRAINING

Learn Data Analytics + Generative AI Together

Our curriculum covers Excel, SQL, Power BI, Python, and Generative AI tools — including prompt engineering for analytics. Live instruction. Real projects. Free demo class.

Book Free Demo →