Generative AI has triggered one of the biggest disruptions to technical hiring in a decade. Candidates can now write production-quality code, craft convincing system design essays, and generate detailed SQL queries — all with AI assistance — in minutes. Indian tech companies from TCS and Infosys to Razorpay and Atlassian are scrambling to redesign their hiring processes in response. This guide explains what is changing, how companies are adapting, and what this means for both candidates and hiring teams.
The Disruption: What Generative AI Enables Candidates to Do
| Task | Before GenAI | With GenAI (2024) |
|---|---|---|
| Take-home coding assignment | 4–8 hours of work | 20–40 minutes with GPT-4 / Claude / Copilot |
| System design write-up | Deep architectural knowledge required | Detailed answer generated from a prompt |
| SQL query writing | Requires SQL fluency | Accurate queries generated instantly |
| Code explanation | Requires understanding the code | AI explains any code in plain English |
| Debugging challenges | Days of troubleshooting experience | AI diagnoses and fixes most bugs in seconds |
| LeetCode / DSA problems | Pattern recognition from practice | Many problems solved directly by AI |
The core problem: Companies cannot tell if the candidate solved the problem or the AI did — at least not without redesigning the process.
How Indian Companies Are Responding
Response 1: Live Coding in Monitored Environments
The most direct response. Companies are moving from take-home assignments to live, monitored sessions.
| Platform | Used By | What It Does |
|---|---|---|
| HackerRank CodeScreen | TCS iON, Wipro, Infosys | Live coding with screen recording, tab switching detection |
| CoderPad | Flipkart, Razorpay, Dunzo | Pair programming-style live interview |
| Codility | Zomato, Amazon India | Timed live assessment, AI-usage detection |
| LambdaTest Interview | Various startups | Browser-based live coding |
| Mettl (Mercer) | BFSI, IT companies | AI-proctored, webcam-monitored assessments |
Response 2: AI-Aware Problem Design
Instead of preventing AI use, some progressive companies are designing problems where AI-generated answers fail.
| Old Problem Type | New Problem Type |
|---|---|
| “Write a binary search function” | “Debug this broken implementation of binary search in our legacy codebase” |
| “Design a URL shortener” | “Our URL shortener is failing at 10K RPS — here’s the specific error log. Diagnose and propose a fix.” |
| “Write a SQL query for X” | “This query is returning wrong results. Find the bug and explain why it’s happening.” |
| Generic system design | Company-specific system design: “Design our next feature given these constraints” |
Context-rich, company-specific problems resist AI solutions because the context is not publicly known.
Response 3: Oral Follow-Up to Any Written Submission
A growing number of Indian companies now ask candidates to explain their code or design submission on a live call.
> “You submitted a solution for our take-home. Walk me through your approach. Why did you choose this data structure? What would you change if the dataset was 100x larger?”
The tell: Candidates who used AI but don’t understand the output struggle to explain their own submission. This is now the primary detection mechanism.
Response 4: Behavioural and System Thinking Deep Dives
Companies are shifting evaluation weight from “can you write code?” to “do you think like an engineer?”
| Old Emphasis | New Emphasis |
|---|---|
| LeetCode algorithmic correctness | Engineering judgement and trade-off analysis |
| Syntax and implementation | Problem decomposition and prioritisation |
| Getting the right answer | Explaining your reasoning process |
| Take-home project output | Live discussion of decisions made |
What This Means for Candidates in India
The honest truth: Using AI to complete take-home assignments or assessments without disclosing it is academic dishonesty. Many Indian IT companies now have explicit AI usage policies in their assessments — violating them can result in immediate disqualification and blacklisting.
The strategic truth: AI fluency is now a skill itself. Companies are beginning to embrace candidates who can use AI tools effectively as part of their workflow — and many assessments now explicitly permit AI tool use.
| Scenario | Right Approach |
|---|---|
| Assessment explicitly says “no AI tools” | Do not use AI — detection has improved significantly |
| Assessment doesn’t mention AI | Ask the recruiter explicitly before using |
| Assessment says “AI tools permitted” | Use them — but be prepared to explain every line |
| Live technical interview | AI is irrelevant here — your thinking is on display |
Skills that AI makes more valuable (not less):
- Code review and debugging: understanding what the AI got wrong
- Architecture and system design: AI generates options, humans choose and justify
- Requirement analysis: translating business needs into technical requirements
- Communication: explaining technical decisions to non-technical stakeholders
How AI Is Being Used in the Hiring Process Itself
Generative AI is not just affecting candidates — it’s being used by hiring teams too.
| AI Hiring Tool | What It Does | Used In India |
|---|---|---|
| Resume screening AI | Ranks and filters resumes by match score | TCS iON, Naukri ML, Keka AI |
| JD generation | Creates job descriptions from role summaries | HRMs like Darwinbox, Keka |
| Interview question generation | Generates role-specific questions | Emerging — not mainstream yet |
| Candidate scoring | Scores assessment responses using NLP | HackerRank, Mettl, Talview |
| Offer letter drafting | AI-drafted offer letters | Some MNCs using internal tools |
| Background check summarisation | Summarises BGV reports | AuthBridge AI layer |
For Hiring Teams: Rethinking the Technical Assessment
Old technical hiring process:
1. Screen resume (ATS)
2. Phone screen (30 min)
3. Take-home coding assignment (4–8 hours)
4. Technical interview (1–2 hours)
5. HR round
→ Vulnerable at Step 3 to AI assistance
New technical hiring process:
1. Screen resume (ATS + skills verification)
2. Short async video intro (10 min)
3. Live coding in monitored environment (45–60 min)
4. Oral follow-up on any take-home (30 min)
5. System design with company context (60 min)
6. Culture / behavioural interview
→ AI is largely neutralised at every step
References:
- HackerRank — State of Software Engineering Report 2024 — https://www.hackerrank.com/research/developer-skills/2024
- GitHub Copilot — Impact on Developer Productivity — https://github.blog/2023-06-27-research-quantifying-github-copilots-impact
- NASSCOM India — AI in Tech Hiring Report 2024 — https://nasscom.in/ai-hiring
- Mercer Mettl — Assessment and Proctoring India — https://mettl.com/resources
- Economic Times India — AI and Technical Hiring 2024 — https://economictimes.indiatimes.com/tech