Shamim Rajani at FAST University FYP Openhouse
Most days, Shamim Rajani sits on the other side of the table, setting timelines, reviewing deliverables, and asking the hard questions. At the FYP Openhouse at FAST University, Karachi, she found herself in a different kind of conversation, one centered on ideas still taking shape and students taking their first steps into the industry.
The drive across the city was not short. But that was exactly the point. In an age where meetings, updates, and feedback can all happen through a screen, there is still something irreplaceable about being present. Walking through project displays, listening to students explain their work, and engaging with them face-to-face creates a connection that no email or virtual call can replicate.
The Wall of Achievements
FAST University is one of Pakistan’s most respected institutions for computer science and software engineering, and the campus reflects that clearly.
The first thing that pulled us to a stop was the Wall of Achievements. Framed photographs and certificates lined it from end to end, each one marking someone who had built something, earned something, and walked out into the world carrying this place with them. We slowed down, looked at the faces, and read the names. And we remembered, all over again, exactly why we had made the drive.
We walked into the hall where the Final Year Project presentations were already running, and the energy in the room hit us immediately — beautiful, productive chaos. Students clustered around their project standees, rehearsing pitches, recalibrating demos, managing the visible tension of people who genuinely care about what they have built. Moving through the space, we caught fragments of conversation at every turn: Agentic AI, system architecture, agile workflows, real-world problem statements. The vocabulary alone told the whole story. These were not students playing in the industry. They were already inside it.
Academia and Industry — Same Table, Finally
Ms. Anam Qureshi, Head of the Department of Software Engineering at FAST University, welcomed us the moment we arrived and set the tone for everything that followed. This was a genuine exchange between teachers and builders, with students at the absolute center of the discussion.
Shamim reviewed three projects that afternoon. Each one went beyond what the brief required, rich with purpose, effort, and a clarity of thinking that pointed to teams who had interrogated their own ideas before presenting them. These students did not build projects. They built arguments — for why their problem mattered, and why their solution was the one worth backing.
A View from the Evaluation: Three Proofs of Pakistan’s Tech Potential
Floodex — Flood Damage Assessment powered by Agentic AI
The Problem
Flooding hits Pakistan hard every single year, displacing millions, wiping out crops, and overwhelming the authorities responsible for coordinating relief. The data exists. The satellite imagery exists. What is missing is a system that converts all of that raw information into something decision-makers can act on, fast enough actually to matter.
The Solution
The Floodex team built exactly that. Drawing on satellite imagery, texture-based feature extraction, and deep learning models, the platform turns raw geospatial data into structured intelligence that authorities can use immediately. The standout feature was the team’s use of Agentic AI: a multi-agent architecture powered by Vision-Language Models that reads natural-language queries, routes tasks to specialized agents, and generates real-time flood-impact reports. A government official can ask which zones face the highest risk, how many people need to be evacuated, or what the crop losses look like in a specific region, and get a clear, plain-language answer without needing a technical interpreter in the room. For a country that absorbs billions in flood damage every few years, Floodex reads less like a student project and more like a tool Pakistan should already be using.
Floodex runs on the same principles we write about in depth on our website. If agentic AI frameworks have been on your mind- how they’re built, how they scale, what makes a multi-agent system actually work- that’s exactly what we break down here.
Sahaara — Pakistan’s First Unified Community Support Platform
The Problem
Anyone who has tried to find a blood donor through social media in Pakistan understands the problem. Requests go unanswered. Information is unverified. Getting timely, trusted help depends almost entirely on who you happen to know. Community support in Pakistan runs informally and inconsistently, and it fails the people who need it most at the worst possible moments.
The Solution
Sahaara addresses this directly as Pakistan’s first unified community support platform, connecting people in need with verified, nearby helpers for blood donations, medical assistance, and everyday tasks. It replaces the scattered, unreliable network of social media appeals with something structured, trustworthy, and built for urgency. What stood out about this team was the empathy behind every design decision. They started from a lived frustration, asked why something this essential is still this hard, and followed that question all the way to a working platform.
InsureLink — Digitizing the Health Insurance Lifecycle
The Problem
Health insurance claims in Pakistan still go through processes from a different era. Hospitals message insurers over WhatsApp. Employers chase approvals through email threads that stretch for days. Paper forms pass slowly between parties while patients wait. The whole lifecycle runs on manual, fragmented workflows that introduce delays and errors at every step, with no clear accountability when things go wrong.
The Solution
InsureLink replaces all of that with a coherent digital pipeline. Built on a three-tier architecture using NestJS, Next.js, and PostgreSQL, the platform digitizes the full claim lifecycle across patients, hospitals, insurers, and corporate employers at once. Real-time synchronization via WebSockets keeps every party aligned as claims move through the system. The result cuts processing time from weeks to hours and removes the dependency on scattered communication channels entirely. The technical depth here was immediately apparent. This team understood the problem at an engineering level and built something that genuinely matches that understanding.
What No One Tells You Before Launch Day
“Judging means recognizing a team’s true focus rather than merely scoring projects: what they chose to solve, how they framed the problem, and whether they have thought beyond the demo.” – Shamim Rajani
Shamim left that afternoon with a strong impression. These students understand technology at a depth that would have surprised industry professionals just a few years ago. They know what Agentic AI is, how multi-agent systems work, and how to talk about it all with genuine confidence. The gap between what they know and what the market expects has never been smaller, and that is worth saying out loud.
But a few honest gaps came up across the afternoon, and they are worth naming because they are exactly the things the industry should be walking into universities to say.
- Equity and ownership. When a team builds something with real startup potential, the question of who owns what needs a clear, documented answer early on. Defining equity upfront protects the team, the idea, and the relationships that made it possible.
- Marketing is not an afterthought. The research was solid. The technical work was solid. What was consistently missing was a credible answer to the question the market asks first: once this product exists, how do people find it? How does a new platform earn trust? The best product in any category still needs a deliberate plan to reach the people it was built for.
- Think like a founder, not just an engineer. If you believe your product solves a real problem, that belief deserves a bigger stage than a university hall. How do you reach the investors who care about exactly this kind of solution? What does the next six months look like, not in the codebase but out in the world? These are the questions that turn a Final Year Project into a first chapter.
What We Carried Back Across the City
Pakistan’s technology potential does not live in the future tense. It lives in university halls, defended by 22-year-olds who stayed up all night to get the demo right. They pursue Agentic AI and real-world problem spaces because they go after it every day. The gap between what they know and what the market expects has never been smaller.
Industry owes them one thing: show up, speak honestly, and take their work seriously. That is what corporate social responsibility looks like at Genetech Solutions. Not a line in a report. A presence. This was not our first visit of this kind. Shamim recently sat on the jury at NED University’s Final Year Project Expo and found the same energy, the same ambition. Read that story here.
We crossed the city for a FYP Openhouse. We came back with a renewed conviction about where Pakistan’s tech industry is actually headed. For more perspectives on Pakistan’s technology landscape and the ideas driving what we build at Genetech Solutions, explore the Genetech Solutions Knowledge Hub.
