Inside Tech Careers: From Tier-2 College to Adobe

Main ideas & lessons conveyed

Career background & role at Adobe

Guest is Sham (SD5 / Senior Computer Scientist) at Adobe. His day-to-day responsibilities include:

• Mentoring junior/new engineers
• Writing code
• Participating in design reviews and code reviews
• Architecting and creating/expanding services

Education & early preparation

• He studied at JPN (mentioned as “JP Na…”, likely in an “Information Technology, Sector 62” context).
• He was a strong student but not a “topper.”
• When he graduated, there were fewer resources and less clarity on interview expectations (including DSA expectations). So his preparation focused on:
  • Ensuring fundamentals were clear from the start
  • Clearing interviews when companies came to campus
• Adobe did not visit his campus, so he gained an off-campus opportunity.

Misconception about Adobe/product companies

Common misconception: engineers only add small features or do patching, because products are mature.

Correction: even early-career engineers can:

• Own services
• Contribute to service design
• Architect and write production code

He observed this in his own team.

Tier-2/3 vs IIT-style access

He argues it’s harder when a company doesn’t visit your college—you must prove yourself through:

• Resume
• Skills
• Projects
• Internships
• Open-source
• DSA
• (and more)

However, he insists it’s possible with the right skills and extra effort.

---

Detailed preparation methodology / roadmap (explicitly given)

Year-by-year roadmap (for a tier-2 student starting over)

Year 1: fundamentals only

• Master core engineering foundations
• Build strong understanding of internals and fundamentals (same for all streams)
• Goal: clear interviews relying on fundamentals

Year 2: DSA + mini projects

• Start Data Structures & Algorithms (DSA) problem solving
• Do mini projects
• DSA is treated as training for:
  • How to think about complex problems
  • Problem decomposition and solution strategy
  • Skill applicable to real engineering tasks (even if not always graphs/trees specifically)

Year 3: internships

• Focus on securing a good internship
• Do advanced DSA alongside internship prep

Year 4: placements

• Focus on whatever is needed for placements
• Re-emphasized interview-relevant fundamentals:
  • OS
  • Networks
  • (and overall core fundamentals typically examined)

Projects vs internships (current guidance)

• Projects still matter, but:
  • With AI-generated projects, it’s harder to verify authorship/effort.
• Internships matter more than projects nowadays, because:
  • They prove exposure to real-world problems
  • They show how to work in an organization

When to apply for internships

• Top colleges (IIT/NIT): companies visit early for internship hiring
• Tier-2/3 colleges: start applying independently by checking:
  • Company pages / careers pages
  • Internship hiring postings
• Prerequisites before applying:
  • Solid resume
  • GitHub with visible DSA practice/graphs
  • At least 1–2 projects you thoroughly understand
• Target timeline:
  • Mid of Year 2: start applying
  • Year 3 summer: do the internship

---

System design & its growing importance

Why system thinking matters for big platforms

Large-scale services must handle massive traffic spikes (example: major live sports events). System design includes thinking about:

• Performance impact
• Load and traffic handling
• Reliability (e.g., “website should never crash”)

Example principle: rate limiting should not introduce unacceptable latency. You must design so protections don’t harm user experience.

Why he teaches system design

He claims many people can write code well, but lack skill in:

• Designing systems
• Architecting services/subsystems

System design teaching provides practical experience from designing scaled systems.

When students should start system design

Earlier he didn’t recommend it, but now—due to interview changes and AI—he suggests starting around Year 3.

Rationale:

• By Year 2, you learn DSA and do projects
• Then system design becomes the next step

What to learn in system design (broad coverage)

• Not “one” system topic only.
• Learn at least one design problem from each category, because:
  • Each category teaches reusable concepts

Categories/examples mentioned:

• Social media (Instagram/Facebook)
• Location-based services (Uber/Ola/Zomato)
• Search systems (Twitter/WhatsApp search, etc.)
• Chat/messaging (WhatsApp/Telegram)

---

AI and software engineering (job risk, skills, workflow)

Will AI replace software engineers?

His view: AI won’t “take jobs”; it augments engineers.

• AI mainly accelerates work and reduces boilerplate/repetitive tasks.
• Deep thinking and responsibility remain with humans.

Who is most at risk?

Engineers who don’t use AI effectively, and therefore don’t get productivity gains.

If the work still takes the same time as before, it becomes “not justified” relative to:

• Salary cost
• AI costs/credits

Emphasis: use AI to reach higher output levels (e.g., “2x/5x/3x” framing).

Roles that become more valuable

• New roles can emerge (example mentioned: FD deployed engineers—framed as “new roles”).
• Quality Engineering (QE/QE) becomes more important because:
  • AI can generate code quickly, but humans must ensure correctness
  • QA thinking and failure understanding is still not fully automated
• He also highlighted:
  • System design/architectural capability
  • Agent/AI-related skills (mentioned as “AI roles” / “architects” being important)

What still matters even with AI

You still must be able to:

• Write and review code
• Understand algorithms and problem-solving

DSA remains important because it builds:

• A problem-solving mindset
• Optimization thinking (reasoning about complexity changes)

How AI has changed his workflow

Previously: manual coding and reviewing, PRD/LLD steps. Now: agents assist across the SDLC:

• Requirements → detailed PRD (agent drafts; he reviews)
• PRD → LLD (agent drafts; he validates/refines)
• Agent writes code and does preliminary work; he reviews and takes over as needed
• On-call debugging steps partially automated by agents/harnesses

Skills to invest in over the next five years

• Code review and debugging become even more important because AI writes code and humans can’t trust it blindly.
• AI fundamentals for engineers:
  • Understand transformer models at least conceptually
  • Know how prompts lead to responses
  • Don’t treat models as a complete black box
• Track/model awareness:
  • Companies release new models/features frequently, and engineers need to stay updated.

---

Rapid-fire Q&A takeaways (key answers)

• Best investment among DSA + system design vs web dev: recommends DSA + system design first.
• Remote vs office: prefers hybrid
  • In-office for collaboration/meetings
  • Home for deep focus/coding time
• Most-used AI assistant: Claude
• Most underrated tech skill: system internals (Linux, memory allocation, OS-level understanding)
  • Cited performance/cost improvements from OS/Linux-level optimization
• Toughest interview: not framed as “tough”; he liked interviews aligned with the company’s real work and discussed it with him (even if not fully solved then).
• Habit that changed his career: staying updated with new industry models/technologies and learning how newer models differ internally and why they’re better.
• Advice that could have saved him years:
  • He started the AI “race” late
  • Earlier adoption of agent development/AI could have saved time
  • SDLC should involve agents doing the first cut of work; he still reviews each step
• Interview success factor:
  • Approach/thinking process matters more than immediately getting the final answer
  • Recommend:
    • Think out loud
    • Show approaches, options, tradeoffs
    • Do regular check-ins with the interviewer
• Is CGPA truly important for placements (off-campus/on-campus)?
  • Only to meet minimum cutoffs
  • Once eligible, selection depends on interview performance, not CGPA
• Closing advice:
  • Don’t worry excessively about layoffs
  • Focus on being prepared for any situation (especially future interviews)

---

Speakers or sources featured

• Nisha (host / interviewer)
• Sham (guest; “Shim/Shikar” in subtitles—same person), Senior Computer Scientist (SD5) at Adobe, and a system design mentor at GeeksforGeeks (as described in the intro)
• Referenced companies/brands/examples (not speakers):
  • Adobe, Netflix, YouTube, Hotstar, Star/Hotstar (mentioned), Google, Instagram, Facebook, Uber, Ola, Zomato
  • Twitter, WhatsApp, Telegram
  • Linux/OS topics, OpenAI, Anthropic/Claude, Gemini
  • Entropic, GeeksforGeeks, IITs, NITs
  • JV/JP College (as written “JP Na…”, “Sector 62” context)