AI Agents | LLMs, Computer Vision & Full-Stack Dev
92
Followers
AI Agents | LLMs, Computer Vision & Full-Stack Dev
🚀 Most AI outreach tools stop at writing messages.
Warmo.ai (http://Warmo.ai) actually helps you find prospects, personalize outreach, and start meaningful conversations—all from one platform.
Whether you're a freelancer, agency owner, recruiter, or founder, it's built to save hours of manual work and help you book more meetings.
🔥 Best part? The first users can use it completely FREE.
If you've been waiting for an AI tool that actually helps you grow your business instead of just generating text, now's the time to try it.
👉 Check out Warmo.ai (http://Warmo.ai) and claim your free access while it's available.
#AI #Sales #LeadGeneration #Outreach #Freelancer #Agency #Startup #BusinessGrowth #Automation #Contra
7
267
Turning long documents into decisions.
Organizations rely on long reports to make critical decisions, but most insights stay buried.
This AI Decision Intelligence system turns raw documents into predictions, evidence scoring, and clear insights in minutes.
If AI could read your reports for you, what decisions would you automate first?
#AI #MachineLearning #DecisionIntelligence #EnterpriseAI #DataAnalytics #Automation
15
463
🚨 Turning CCTV Into a Crime-Prevention AI System
We helped build Aegis FaceGuard, an AI-powered security platform designed to prevent armed robberies before they happen—not after.
In high-risk regions like Colombia, traditional CCTV only records.
This system predicts, detects, and reacts — instantly.
Here’s what it catches in real time:
🪖 Helmet detection
🔫 Gun detection
🎠Mask detection
👤 Suspicious posture
⚠️ Threat movement patterns
When danger appears, the system automatically:
Triggers alarms
Alerts police
Notifies the owner
Saves evidence — all within milliseconds
This is security before the crime, not after.
If you're building AI for public safety, retail protection, or real-world automation, I’d love to connect. 🔥
#AI #ComputerVision #SecurityTech
15
313
AI Legal Case Analysis — Expert Evidence Prediction
This project demonstrates an AI tool that analyzes legal cases and predicts how a judge may view expert testimony. Using structured case data and LLM reasoning, the system compares past rulings involving Dr. Bailey with a new case scenario.
By parsing the actual judgment text:
“I accept Dr Bailey’s evidence…”
The model evaluates patterns in credibility, cross-examination outcomes, and judicial language. It then generates a predicted ruling, concluding that the judge would likely consider Dr. Bailey’s expert evidence credible again.
This showcases:
Smart case-to-case comparison
Expert-evidence credibility scoring
LLM legal reasoning
Structured JSON analysis
#LegalTech #AIforLaw #CaseAnalysis #LLMReasoning #ContraPortfolio
Spent the last few months rebuilding the digital presence for Shiva Group, a multi-brand automotive dealership group running JLR, Mahindra, Triumph, KTM, Bajaj, and MG Select showrooms across North India.
New site architecture, cleaner model pages, a live outlet locator across 50+ locations built entirely on Framer. Genuinely happy with how it turned out, and it's already paying off for them too, with $1.2M+ in vehicle bookings coming through the site in the first month!
Check it out (https://contra.com/p/Ekbr6eEZ-shiva-group-full-website-rebuild-across-10-automotive-brands)
Muddy Paws: a booking engine that turns "can I book?" into "you're booked" in 60 seconds
Business: Muddy Paws Mobile Grooming, Lisbon, Portugal as Early Stage !
One-line problem solved: A solo mobile dog groomer was losing bookings and about 1 in 5 appointments to no-shows, so I built a booking flow that confirms and protects every slot with no manual back-and-forth.
The client brief:
InĂŞs runs a grooming van in Lisbon, alone, booking about 6 dogs a day, Tuesday to Saturday.
Requests arrive through WhatsApp and Instagram DMs and get lost.
She can't answer the phone while grooming.
About 1 in 5 customers don't show up, which costs her the slot and the fuel.
Appointment length depends on the dog's size and coat.
She drives between neighborhoods, so not every time slot works.
The goal: turn an inbound inquiry into a confirmed, paid booking with as little effort from InĂŞs as possible.
Before and after
Before: a DM arrives, InĂŞs replies hours later, they negotiate times, the customer forgets, the dog doesn't show, and the slot is gone.
After: the customer picks their dog and services, sees only the slots InĂŞs can actually serve, pays a small deposit, and gets an instant confirmation. Reminders, rescheduling, and waitlist refills happen automatically.
What I built:
Fixing the front door
Dog profile and service selection, with duration and price calculated live from size and coat.
Address check against InĂŞs's service zones.
Route-aware availability: slots respect her neighborhood-by-day schedule, service length, and travel buffer.
Deposit to secure the slot, then an instant confirmation with calendar links.
Guest checkout with no account needed.
Fixing the follow-through:
Reminders at 48 hours and 2 hours with one-tap confirm, reschedule, or cancel.
Cancelled slots reopen and go to a waitlist automatically.
A nudge for customers who start a booking and don't finish.
An owner dashboard: today's route, a "Needs attention" queue, and hours saved.
Built to work worldwide: currency, timezone, language (including Arabic right-to-left), units, and service zones are configurable, so the same engine can serve a groomer in Lisbon or a dentist in Toronto.
How it maps to the criteria:
Problem-solving impact: a complete path from inquiry to a paid, confirmed booking.
Owner-effort reduction: confirmations, reminders, rescheduling, waitlist, and lead follow-up all run without InĂŞs.
Craft and execution: mobile-first, branded, and tested end to end with seeded demo data.
Storytelling: the demo video moves from a chaotic WhatsApp day to a calm dashboard with the hours-saved number.
Process:
I treated it like a real client brief: define the owner's quirks first, then design the booking logic around them (duration, route, deposits), then build in Lovable in layers, starting with the booking flow and then automations, dashboard, and global settings.
I’m a little late to the party with this one, but when you’re building a living, breathing AI with a short fuse, you don’t rush the process.
Designers, we’ve all been there: The logic is flawless. The grid is perfect. Then... the feedback arrives. "Can we make it more... blue-ish?"
Inner Designer is my latest experimental project that says what you can't. Built with Figma Make, a live backend, and an AI brain that actually understands design theory, it doesn’t just design—it defends.
The Hook: It explains its choices.
The Twist: It has a breaking point.
The Result: Push it too far with bad revisions, and the UI literally melts into frustration.
It’s part tool, part therapy, and a little too real.
Here is the Live Link: Inner Designer (https://moon-morph-57934661.figma.site)