Two Internships, Same Final Semester
I'm finishing my B.Tech in Computer Science Engineering at Aravali College of Engineering and Management, and over the course of my final year I've worked two internships that, on paper, look adjacent but actually taught me very different things: AI Intern at Draskenlabs, then RPA Developer Intern at Avent IQ. Neither was a research role or a big-name product team — they were the kind of internships most CS students actually get, and I think that's exactly why the lessons are worth writing down.
What the AI Intern Role Actually Involved
At Draskenlabs, the day-to-day wasn't "build an AI model." It was:
- Assisting with frontend development and customization for CRM-based web apps
- UI improvements and form handling — unglamorous but constant work
- Helping develop workflow solutions using AI and automation tools
- Building an automation workflow using n8n to streamline task execution
The gap between "AI Intern" as a job title and what I actually spent time on was instructive. Most of the AI work wasn't training or prompting anything exotic — it was figuring out where an LLM node could replace a manual decision inside an existing workflow, and making sure the rest of the pipeline (the form handling, the CRM updates) was solid enough that the AI piece wasn't sitting on top of a shaky foundation.
What the RPA Developer Intern Role Actually Involves
At Avent IQ, the work is more structured and closer to what "RPA Developer" sounds like:
- Developing and maintaining RPA bots for business process automation
- Monitoring bot execution and validating workflows in production
- Testing bots, debugging issues, and resolving them
- Creating and maintaining Process Definition Documents (PDDs)
The single biggest adjustment moving from Draskenlabs to Avent IQ was the shift from "does this workflow produce a reasonable output" to "does this bot produce the exact same output every single time, and if not, why not." RPA work rewards precision and documentation discipline in a way that AI-workflow work doesn't demand as strictly — a PDD has to describe the process accurately enough that someone else could debug the bot without asking me a single question.
Lesson 1: Certifications Were Useful, But Only Alongside Real Work
I hold an n8n Level 1 Certification, an Automation Anywhere Essential Certificate, IBM SkillsBuild's Introduction to Virtual Intelligent Agents, and Anthropic Academy's Claude Code in Action. None of them, on their own, taught me what actually shows up in production automation work. What they did was give me vocabulary and a mental model fast enough that I wasn't learning the tool and the job simultaneously — by the time I was debugging a bot in Automation Anywhere at Avent IQ, I already understood the platform's building blocks and could focus on the actual process logic instead of the UI.
My honest advice: treat certifications as a way to skip the "what does this button do" phase, not as a credential that replaces hands-on work. The n8n certification made my first week at Draskenlabs faster, not more impressive.
Lesson 2: RPA Discipline Made Me Better at AI Automation, Not Worse
I expected the RPA internship to feel like a step backward after doing "AI" work — rule-based bots seemed less interesting than LLM-driven workflows. In practice, writing PDDs and monitoring bot execution in production taught me to think precisely about failure modes, edge cases, and documentation in a way that directly improved the AI/n8n workflows I build outside of work, including personal projects. An LLM-based workflow without that same discipline — clear failure branches, a record of what the automation is supposed to do, monitoring instead of "set it and forget it" — tends to degrade quietly until something breaks in a way nobody can explain.
Lesson 3: Being Selected for the Infosys Program Was a Signal, Not a Shortcut
I was selected for the Infosys IT Training Program in Data Analytics, which mattered less for what it directly taught me and more as a confirmation that the direction I was investing in — data, automation, structured problem solving — was one worth continuing. If you're a student stacking certifications and program selections, my honest take is: use them as checkpoints to validate direction, not as the goal itself. The internships are where the actual skill compounds.
Lesson 4: The Overlap Between Roles Is Bigger Than the Job Titles Suggest
"AI Intern" and "RPA Developer Intern" sound like different tracks, but the overlapping skill underneath both is the same: breaking a business process into an explicit sequence of steps, deciding which steps are deterministic and which require judgment, and building something that runs reliably without me babysitting it. Automation Anywhere and n8n are different tools for expressing that same underlying skill. If you're a student trying to decide between an "RPA" internship and an "AI/automation" internship, my advice is: the job title matters less than whether the role actually gives you ownership over a real process end-to-end, from documentation to debugging.
Advice for CS Students Considering This Path
A few concrete things I'd tell a younger version of myself starting this path:
- Get one certification in a rule-based tool (RPA) and one in a workflow tool (n8n) before you need them. Having both vocabularies makes you useful in either kind of internship, and lets you have an informed opinion on which fits a given problem.
- Ask for the PDD or process documentation, even if nobody assigns it to you. Writing down exactly what a process does is the fastest way to actually understand it, and it's a skill that transfers to every automation job regardless of the specific tool.
- Don't treat "AI automation" as strictly better than "RPA." I've made this mistake myself. The rule-based bot that never needs an API call and never hallucinates is sometimes the more correct engineering choice, and understanding when that's true makes you a better automation engineer, not a less impressive one.
- Build something of your own alongside the internship work. My personal projects (a Chrome extension paired with an n8n workflow, an AI-powered fitness app, a Python-based face recognition pipeline) gave me a space to make architecture decisions nobody was making for me, which is exactly the muscle an internship alone doesn't always build.
Where I'm Headed
My RPA Developer Internship at Avent IQ runs through July 2026, right alongside finishing my final semester. I don't think the next step is choosing "RPA" or "AI" as a permanent lane — the more useful skill, based on what these two internships have actually taught me, is knowing which one a given problem actually calls for.
Reach Out
If you're a student navigating similar internship decisions, or want to talk RPA, n8n, or automation careers generally, I'm happy to compare notes. Reach me at rishabnishad22@gmail.com, on WhatsApp, or through my contact page.