Innovative Projects
Explore my cutting-edge projects in AI and computer vision.
AI Solutions Development BITS Pilani
Developed an AI-driven analytical tool at BITS Pilani designed to evaluate students’ interests and interaction patterns with different laboratory machines. The system leverages CCTV footage and computer vision algorithms to track individual student engagement levels with specific equipment. Using deep learning-based face and activity recognition, it identifies which machines attract the most attention from each student and quantifies their duration and frequency of usage.
In addition to measuring engagement, the tool automatically logs lab utilization patterns, provides detailed reports on overall machine usage time, and maintains individual attendance records. The resulting data insights enable faculty to understand students’ practical preferences, optimize lab scheduling, and enhance resource allocation efficiency.
Computer Vision
Something interesting is happening in restaurants.
For years, CCTV cameras sat in the corner — always watching, rarely used. Just silent witnesses… until a theft, a fight, or a complaint.
But now? The smartest restaurant operators are rewinding the tape — and fast-forwarding their business.
After months of research and conversations with leading restaurant brands across India and beyond, I uncovered a shift:
CCTV is no longer just about security — it’s becoming a secret weapon for operational excellence. Here’s what I discovered from intelligent CCTV systems:
🔍 Customer behavior: Spot patterns that influence everything from table layouts to menu design.
👥 Staff performance: Track SOP adherence, service quality, and peak-hour response — without guesswork.
🔥 Kitchen efficiency: Detect bottlenecks, wasted movements, and prep slowdowns that silently drain margins.
🛑 Loss prevention: Go beyond theft — uncover product wastage, process leaks, and internal slippage.
It’s not about surveillance. It’s about clarity.
And the smartest brands are already turning footage into feedback, and feedback into growth. Cameras aren’t just watching anymore. They’re working. Are yours?
Work Experience, What have I done in IIT Madras?...
I've helped create facial recognition so cheap it costs less to run than buying a kilogram of apples.
Our team built a pose-invariant system that deploys instantly on any CCTV network, slashing operating costs by 100X through deep learning optimization.
But now I see the implications.
Yes, there are security benefits - only authorized access to sensitive areas, safer spaces for our loved ones.
But watching our system identify people with frightening accuracy in our deployments across 200+ identities in offices has been sobering.
This isn't new technology - DeepFace beat human-level recognition back in 2014. But seeing it deployed at this scale and cost is something else entirely.
As technologists, we're only now grasping our weight in shaping society.
We created something powerful, but at what cost to privacy? To freedom of movement without constant identification?
Have we crossed a line between security and surveillance?
Innovative Solutions in AI and Computer Vision
For years, CCTV cameras in restaurants were like unused gym memberships—always running in the background, rarely actually used. They were there “just in case”: for theft investigations, a customer dispute, or a hygiene issue flagged by an inspector.
But day-to-day operations? Cameras sat idle, collecting hours of footage no one watched. They were treated as insurance, not intelligence.
That’s Changing—Fast.
Today, restaurant chains in India, Southeast Asia, and the Middle East are waking up to a powerful idea: What if cameras could help run the business, not just record it?
Thanks to AI and video analytics, cameras are no longer passive observers. They’re becoming active participants in operations.
They can spot long queues before customers walk away.
They flag empty counters when no staff is around.
They identify missed clean-ups and track hygiene compliance in real time.
Instead of watching hours of footage after something goes wrong, restaurants are now getting live alerts and actionable insights. What used to be just video is now real-time operational data.
Why Now?
Running a single restaurant is hard. Running 50 is chaos—unless you have systems to maintain consistency. Managers can’t be everywhere.
Feedback gets filtered.
Audits are infrequent and often too late.
POS systems tell you what was sold, but not how the service experience went.
CCTV with AI changes that. It becomes a tool that watches every outlet, every shift, 24/7—with zero bias and full memory.
It’s Like Giving Every Outlet a Smart Assistant
AI video tools don’t just record. They analyze. They understand the context of what’s happening and surface only what matters:
“This counter has been unattended for 8 minutes.”
“That employee skipped part of the hygiene routine.”
“A queue formed and no one responded for 3 minutes.”
It’s not just about saying something went wrong—it’s about showing what, where, and when, so teams can fix it fast.
So, Why Is This Happening Now?
Several forces are coming together:
Rapid Scaling, Limited Ops Maturity India’s QSR (Quick Service Restaurant) market is growing at over 23% a year. Brands are expanding faster than their operations can keep up. You can't scale chaos.
Human Monitoring Doesn’t Scale More locations mean more complexity. Field visits, audits, and human checks aren’t enough. Staff turnover makes training and consistency harder. You need a system that works even when people don’t.
Hygiene Is Non-Negotiable Post-COVID Customers don’t just expect cleanliness—they expect visible, provable hygiene. Your CCTV can now demonstrate compliance and build trust.
Tech Is Finally Affordable and Scalable What once required expensive servers and custom hardware now runs on existing cameras with cloud or edge AI. This makes advanced analytics accessible even to mid-sized chains.
AI Saves Money by Preventing Mistakes Labor costs are rising. Operational errors are costlier than ever. AI helps managers focus on real issues, not gut feelings or guesswork.
What the Future Looks Like
This shift is more than a tech upgrade—it’s a paradigm shift. CCTV + AI is becoming the nervous system of restaurant operations.
It’s not about watching people. It’s about understanding the environment. It gives restaurant brands the ability to scale with control, standardize quality, and respond instantly to issues—no matter how big they grow.
Anitum: Making This Future Possible
At Anitum, we help restaurant chains turn their existing CCTV footage into a daily operations tool. We extract insights, patterns, and alerts—so your ops team doesn’t have to guess or play catch-up. They can act, adjust, and improve—before problems affect customers.
Because video shouldn’t just tell you what went wrong. It should help make sure things go right.
“Restaurant cameras are like gym memberships—always on, rarely used.”
CCTV is underutilized—only checked post-incident
Ops teams rely on audits, guesswork, or hearsay
Inconsistent service quality across locations
Managers can’t be everywhere
Growing hygiene and service expectations post-COVID Why It Matters
Customer Experience Drops when stores scale
Operational Blind Spots = Lost Revenue
Manual audits are infrequent and expensive
67% of diners say hygiene impacts repeat visits (Deloitte)
Staff turnover disrupts training and consistency
The Shift Is Here
“What used to be passive video is now live data.”
Across India, SEA, and the Middle East, restaurants are:
✅ Detecting long queues in real-time
✅ Monitoring hygiene checklist completion
✅ Spotting empty counters and idle staff
✅ Ensuring SOPs are followed across all outlets
Anitum’s Solution
🎯 AI-Powered Video Analytics for Restaurants
We analyze existing CCTV footage to provide:
Real-time alerts (queues, hygiene, idle counters)
SOP monitoring (staff behavior, table clearing, kitchen flow)
Automated reports and heatmaps
Benchmarking across locations
Cameras in Restaurants: From Passive Watchers to Smart Assistants
Get in Touch
I'm passionate about innovation. Let's collaborate on advanced AI and computer vision projects together!