Video summary

From Hindi Medium to Cracking Big Tech | Saurabh Agarwal | Software Engineer Roadmap | Referral | AI

Main summary

Key takeaways

News and Commentary

Saurabh Agarwal: Big Tech Transition Roadmap (Hiring + AI Era)

Saurabh Agarwal (upcoming Principal Software Engineer at Apple, San Francisco; ~12 years’ experience across Samsung, Amazon, and startups) shares a detailed roadmap for breaking into and transitioning within big tech—especially under current conditions where AI adoption is rising but hiring is tougher.

He argues that success comes less from “shortcuts” and more from:

  • sustained preparation,
  • smart interview execution, and
  • a high-efficiency job search system—especially referrals, LinkedIn activity, and quantity of applications.

Career Journey & Key Context

  • He describes starting from Hindi-medium education, moving through tier-focused colleges, and reaching major employers through competitive recruiting (including mass hiring pathways like Samsung/TCS-Infosys-style routes).
  • At Amazon, he claims experience across multiple teams and working on/building/reviewing major systems (including healthcare and broader platform work).
  • He moved to startups to learn tooling and “how startups work,” then joined Apple for career progression.
  • He emphasizes the hiring shift: AI is booming, but employee demand has decreased relative to earlier cycles—so candidates need stronger, more updated skills and better interview strategy.

Interview Preparation: What to Focus On (and How)

He recommends layered preparation depending on experience level.

For Freshers / Early Career

  • DSA (Data Structures & Algorithms) should dominate (he claims ~80% of focus).
  • Prepare using common curated question sets (he repeatedly mentions lists like Blind 75 / Blind 150).
  • He emphasizes a strategy covering logic + ethics + design fundamentals + behavior.
  • For system design, he recommends learning basic concepts (e.g., parking lot-style systems) rather than deep LLD for large companies.

For Seniors / Staff+ Level

  • Behavior becomes the most difficult and most decisive interview component.
  • He warns that “behavior questions” (e.g., “Tell me about yourself”) can lead to rejection if answered:
    • wrongly, or
    • too literally (e.g., giving personal details that don’t fit the role narrative).
  • System design still matters, but the balance shifts:
    • behavior takes a larger share,
    • system design becomes more supporting.

AI Readiness and “RAG” Basics

  • He claims almost nobody truly understands AI deeply, but candidates can still interview effectively by mastering core vocabulary and concepts.
  • Key interviewable terms include:
    • RAG
    • vector databases
    • LLMs
    • token/usage basics
    • being able to explain RAG at an interview level (e.g., how documents/search results are retrieved and used).
  • If an interviewer goes deeper, he advises you to respond honestly and clearly rather than pretending—avoid over-claiming.

Project & Resume Strategy (Truthful but Strategic)

He prioritizes resume content around:

  • the project section (how projects are presented, including numbers/impact),
  • LinkedIn/resume/design details, and
  • no lying—he explicitly warns that interviewers may “grill” unfamiliar technologies.

For startups specifically, he believes project depth helps because role scarcity forces candidates to demonstrate hands-on building.

Execution During Interviews: Tactics That Prevent Failure

Drawing from interviewing hundreds of candidates, he recommends:

  • Don’t over-assume constraints Ask clarification questions (e.g., whether an array is sorted, how to treat null/empty cases). He emphasizes that assumptions without clarification can cause rejection.

  • Start with brute force first (even if an optimized solution exists)

    • Prove you can solve and decompose the problem.
    • Share optimized steps only after the brute approach is correct and understood.

  • If you’re unsure, use pseudo-code and show your approach The goal is to stay on track rather than freezing.

“Volume Matters” in Job Search

A major theme: candidates often fail due to insufficient application volume or inefficient systems.

  • He claims he applied to very large numbers (e.g., 120 applications to Microsoft with referrals) while expecting only a low conversion rate (calls/offers).
  • His stance: for average engineers, success often requires many attempts, not perfect targeting.
  • He suggests using AI tools to speed up applications without reducing prep quality.

LinkedIn + Referrals: How to Get Interviews

He strongly promotes LinkedIn as the primary lever:

  • Be active: small profile changes, post monthly, comment/like regularly.
  • Message recruiters using a specific “one-message” format:
    • who you are,
    • what you want,
    • what you’ve done,
    • attach your resume.

Referral philosophy

  • Be “a little shameless”:
    • reach out widely instead of waiting to be invited.
  • Ask for referrals from the right level:
    • match your target role level; avoid mismatched-level referrals.
  • If you’re junior and not at the right level:
    • avoid forcing referrals—apply directly.
  • He claims his community drives strong referral activity because it’s reciprocal and active (help job seekers without charging).

Community & Anti-Fraud Stance

  • He describes building a free community for job seekers.
  • He runs free mock interviews and free system-design mock sessions.
  • He strongly warns against paying for resume/interview services marketed on Instagram/elsewhere, including:
    • claims that some “tech trainers” may be fake or outsourced,
    • discouraging paid “ATS resume” scams or shortcuts,
    • arguing that free templates, AI-assisted ATS improvements, and community feedback are sufficient.

Big-Tech Skepticism About AI-Generated Shortcuts

  • He argues that while AI tools (e.g., Claude or ChatGPT) can produce code, companies may still want juniors to code and test themselves.
  • He suggests using AI as a search/assist tool:
    • to understand context and help debug,
    • not to fully delegate engineering work—because code quality and security risks remain.

Educational Background: Don’t Down-Level Yourself

  • He repeatedly tells students not to let tier/degree discourage them (B.Sc/BCA/B.Tech, etc.).
  • His stance: entering the industry matters more than the college brand; once inside, you can upskill and switch.
  • He highlights alternative paths and perseverance, including:
    • QA short courses,
    • GATE/other routes,
    • affordable programs/courses,
    • even US masters later.

Presenters / Contributors

  • Saurabh Agarwal (main speaker)
  • Interviewer/Host (unidentified; asks questions and facilitates the discussion)

Original video