An expert lecture at TKIET on getting a dream job

My own college is TKIET Warananagar, and I have now been there in three different roles.

1. 2011In August, in the first semester of my final year, TCS selected me in a campus placement.
2. 2023On 10 March, I went back with Prudour to recruit. We held a campus placement drive for students from five branches: computer science, mechanical, electronics and telecommunication, civil and chemical.
3. 2026On Thursday, 29 October, I went back to give an expert lecture.

The audience was third- and final-year BE Computer Science students, at Session 4 of the department’s Future Ready Engineers Program 2026-27, titled “How to Get a Dream Job Using Advanced AI Techniques.” Dr. R. B. Patil Sir, Head of Computer Science and Engineering, and Prof. D. B. Mirajkar Sir invited me as the guest of honor.

Have that fear of being average

One thing I want every student in that room to carry with them: Have that fear of being average! Don’t be just another computer engineer.

When a company wants to select only one person, ask yourself, “Why will it choose me and not someone else?” If you can’t answer that yet, finding the answer is your work for this year. AI can help you prepare it. It can’t give it to you.

Downloads

• Student handout (PDF, 1.1 MB): every prompt, command and link from the session.
• Slides as presented (PDF, 3.6 MB): all 30 slides from the day.
• India Fresher Pack and career-ops, both free on GitHub.

A prompt answers once

Every student in that room knows how to ask AI a question. Within one chat, AI remembers what you told it. Its memory is not yet perfect, though: open a new chat, and it may not know who you are, what you studied or which job you want.

If the students remember only one line from the day, I want it to be this: “A prompt answers once. A system works every time.”

A system has three parts. Custom instructions that tell the AI who you are. A second brain: your resume, your project stories and the job descriptions you want, kept as files it can read. And one project for each task you repeat, like interview practice or resume tailoring.

I build every system the same way. I plan it with Claude, stress-test the plan with ChatGPT, get a second opinion from Gemini, and then build it with Claude.

Let an agent do the searching

For the job search itself, I showed three ways, each one a step further than the last.

1. career-ops, on your own computer. career-ops is an open-source job-search agent created by Santiago Fernández de Valderrama Aparicio and built with its contributors. It finds roles, scores each one against your resume and tailors the resume for it. The catch: it runs on your machine, so your laptop has to stay on.

   Indian jobs have their own traps: a CTC that is not your in-hand pay, bonds, a joining date that clashes with your final exams, CGPA and backlog cut-offs. So I made a free add-on that checks for them, the India Fresher Pack.

2. A workflow in the cloud. A Make.com or n8n workflow runs the same search on a schedule, once a day, even with your laptop off. The student handout has a Make.com version for the free plan, based on Abhijay Arora Vuyyuru’s n8n workflow, “AI Agent That Job Hunts While You Sleep.”

3. An always-on AI agent. This is the latest and most advanced way. Personal agents like Meta’s Muse, OpenAI’s Dots, xAI’s Grok Bot and Anthropic’s Claude Duet each get their own computer in the cloud, and they keep working after you close the app. Tell one once: “Every night, find new fresher roles that match my resume, score them and tailor my resume for the top three.” It works while you sleep, and the shortlist is ready in the morning.

Whichever way you choose, one rule stays the same: AI never applies for you. You read what it made, and you press Submit.

career-ops and the Make.com workflow work on free plans. career-ops needs an AI coding tool, and Claude Code needs a paid Claude plan, so it runs free through Antigravity CLI instead. The always-on agents are newer: most of them need a paid plan, and some are available only in the US so far.

The hot seat

A student volunteered for a live mock interview in AI voice chat. The AI asked the questions, she answered out loud, and the AI rated each answer. That is practice anyone can repeat at home, every day, for free.

I left them with one rule. Never use AI during an online assessment or a live interview. AI does the prep. The interview is all you.

Felicitated

They felicitated me with a shawl, a coconut and a specially made trophy. Pramod Pawar and Pavan Lipare, who came with me, were felicitated with saplings.

Earlier that day, we visited the AICTE IDEA Lab with Prof. B. R. Bagane Sir, who coordinates it.

Thank you

Thank you to Dr. R. B. Patil Sir for the invitation and the felicitation, to Prof. D. B. Mirajkar Sir for the invitation, to all the staff for their efforts, and to the students for their full attention and for taking the initiative in the mock interview.