Starting from Zero Practical Experience
I'm a B.Tech Computer Science student at Aravali College of Engineering and Management (YMCA), and like most people in my position, I spent my first couple of years mostly doing coursework — assignments, exams, the occasional class project. None of that translates directly into "I can build something a company will actually use." The gap between classroom CS and being useful to a real team is bigger than most syllabi admit.
What eventually worked for me was a combination of building real (if small) projects on my own, picking up certifications in a specific niche, and being honest in interviews about what I actually knew versus what I was still learning. Here's what that looked like concretely.
Building Projects That Aren't Just Tutorials
The turning point for me wasn't finishing another course — it was building things nobody assigned me. Three projects mattered most:
AI JobCopilot — a Chrome extension that extracts a LinkedIn job description, triggers an n8n workflow, and uses an LLM to generate a personalized recruiter email and an ATS-optimized LaTeX resume for that specific job, then sends it via the Gmail API and logs the application in Google Sheets. I also wired up Telegram so I could submit a job description and get status notifications on my phone. This project forced me to actually learn how automation platforms, LLM prompting, and real APIs (Gmail, Telegram) fit together — not just how to call an API in isolation.
Elevate Fitness — an AI-powered fitness app built with Next.js, Supabase, ShadCN, and Tailwind CSS, with an Agentic AI layer (using MCP servers) that generates personalized workout and diet plans, plus interactive SVG exercise demos, a BMI calculator, and a calorie counter. It's live at elevate-your-fitness.vercel.app. This one taught me how to structure a full-stack app end-to-end, not just a script.
Automated Image Filtering Using Face Recognition — a Python pipeline using OpenCV and InsightFace that detects and matches faces against reference profiles and automatically sorts images into output folders, structured like an RPA pipeline: input, processing, decision-making, output. This project is what actually connected my interest in Python/CV work to the RPA mindset that later became relevant for my internship.
None of these are massive systems. What mattered is that each one was a complete, working thing I could explain end-to-end in an interview — not a half-finished course assignment.
Certifications That Actually Signaled Something
I didn't collect certifications randomly — I picked ones that mapped directly onto the automation/AI niche I wanted to work in:
- n8n Level 1 Certification (n8n Academy) — validated the workflow automation skills I was already using in my projects
- Automation Anywhere Essential Certificate (Automation Anywhere) — directly relevant once I started interviewing for RPA-focused roles
- Introduction to Virtual Intelligent Agents (IBM SkillsBuild) — broadened my understanding of conversational/agentic AI concepts
- Claude Code in Action (Anthropic Academy) — sharpened how I use AI coding tools in my actual workflow, not just as a novelty
I was also selected for the Infosys IT Training Program in Data Analytics, which reinforced the SQL, Excel, and Power BI side of my skill set — the "make sense of data" half of automation work that's easy to overlook if you only focus on building bots.
The Internships Themselves
My first real industry role was as an AI Intern at Draskenlabs (Oct–Dec 2025), where I worked on frontend development and customization for CRM-based web apps — UI improvements, form handling — and assisted in developing intelligent workflow solutions using AI and automation tools, including building an automation workflow with n8n to streamline task execution.
That experience directly led into my current role as an RPA Developer Intern at Avent IQ (Jan–Jul 2026), where I develop and maintain RPA bots for business process automation, monitor bot execution and validate workflows in production, handle bot testing and debugging, and write Process Definition Documents (PDDs).
Looking back, the through-line is clear: personal projects in automation and AI → certifications that backed up those skills on paper → an internship that let me apply it to a CRM/web context → a second internship that let me go deeper into RPA specifically. Each step used the previous one as evidence.
What I'd Tell Someone Starting Where I Started
Build something that solves your own problem first. AI JobCopilot exists because I was tired of manually tailoring resumes and emails for every job application. The best portfolio projects usually start as "I'm annoyed by X" rather than "this would look good on a resume."
Pick a niche and go one level deeper than surface familiarity. Anyone can say "I know Python." Fewer people can explain why they structured a face-recognition pipeline as input → processing → decision → output, and what tradeoffs that structure creates. Depth in one specific area beats breadth across many.
Certifications work best as confirmation, not foundation. I earned certifications for skills I was already applying in projects, not the other way around. A certificate backing up a real project is credible; a certificate with nothing behind it is just a line on a resume.
Be upfront about your actual experience level in interviews. I'm a student. I don't pretend otherwise. What I can speak to confidently is exactly what I've built and exactly what my internships have had me do — and that specificity matters more than vague confidence.
Documentation habits transfer. Writing PDDs at Avent IQ made me realize how much clearer my personal projects became once I started documenting inputs, expected outputs, and failure modes for my own code too — a habit I wish I'd started earlier.
If You're Hiring, Collaborating, or Just Curious
If you're a student figuring out a similar path, or you're looking for someone to help with an automation, RPA, or AI workflow project, I'd be glad to talk. Reach me at rishabnishad22@gmail.com or on WhatsApp, or take a look at my work at rishab-nishad.vercel.app.