Lukáš Caletka
Vue.js and Node.js developer building scalable web apps
New to Contra
Lukáš is ready for their next project!
I created a short cinematic story about a freelance developer who spent hundreds of hours building useful tools, web apps, dashboards, API integrations and product ideas — but kept ending up with unpaid requests, unfinished opportunities and zero paid invoices.
The twist is not that he suddenly becomes happy. The real twist is that he finally turns his work into a real product, gets his first paid project approved, and sees the value of his work come back as income.
Helping people matters, but creators cannot survive by helping everyone for free.
Creative work is still work. Built with Renoise for #RenoiseChallenge.
1
0
143
TrustFlow — an AI scanner and marketplace safety layer for more trusted online transactions
Thanks to the Makeathon and the 10,000 Figma credits I received to continue developing the project, I was able to complete a second, significantly expanded version of the TrustFlow prototype.
The original idea was simple: build an AI tool that helps buyers check risky marketplace items before making a purchase. During the next iteration, the project evolved much further — from a simple AI scanner into a complete SaaS demo for protecting buyers, sellers, and marketplace platforms.
TrustFlow now demonstrates a full safety workflow:
• buyers can upload product photos, videos, or listing URLs
• AI scans the item, visual authenticity signals, seller risk, and suspicious patterns
• the system generates an evidence report with risk scores, AI confidence, and visual markers
• reports can be saved into an evidence vault, shared with sellers, or used during disputes
• sellers can provide missing documentation through a seller review portal
• marketplace operators can review suspicious cases in an admin dashboard
• AI can help create safer marketplace listings from scan evidence, including a product title, description, condition summary, price estimate, authenticity warning, and protected payment recommendation
• listings can include a TrustFlow Evidence Certificate with a clear Needs Review status, missing documentation, escrow recommendation, and a public verification link
A key part of TrustFlow is transparency. The system does not falsely claim that an item is automatically authentic. Instead of using unsafe language like “Certified Authentic,” TrustFlow uses an evidence-based approach: it shows what the AI detected, what documentation is missing, what the seller still needs to provide, and when protected checkout or escrow is recommended.
This new version of the prototype includes:
• a complete AI Product Scanner Demo
• a SaaS App Demo Shell with 6 areas: Scanner, Report, Vault, Seller Portal, Admin, and Billing
• an AI Verified Listing Builder for creating safer marketplace listings
• a TrustFlow Evidence Certificate with careful and transparent wording
• a Public Certificate Verification page with a QR/link preview, audit trail, and 30-day link validity
• a U.S.-style Marketplace Safety Controls strip inspired by seller transparency, review integrity, and buyer-safe disclosure patterns
• a Marketplace Plugin Preview showing how a TrustFlow badge could appear directly inside a marketplace listing
• demo-safe wording that avoids claims of certified authenticity, legal authentication, real escrow custody, live banking, or regulatory certification
TrustFlow addresses several real marketplace problems at once: counterfeit goods, risky sellers, scam messages, off-platform payments, missing documentation, unclear evidence during disputes, and lack of transparency between buyers, sellers, and platforms.
The goal was not just to create a polished landing page. The goal was to show how AI and Figma Make can help move an idea quickly toward a complete product prototype — from the first product scan to a shared evidence report, seller review workflow, and marketplace integration.
TrustFlow is built around one simple idea:
Every marketplace transaction should become safer before the buyer sends money.
Published prototype:
https://radio-ranch-95224118.figma.site
5
334
Flixy Video Editor – Full SaaS Publishing Flow
0
2
Trezzo Canvas E-book Editor – End-to-End SaaS Publishing Flow
0
2
ProofGuard AI – Crypto Security Prototype
0
14
ProofGuard AI – A Crypto Security Layer Against Scams
ProofGuard AI is a prototype of a crypto security tool built in Figma Make. The goal of the project is to act like an “Avast for crypto” — a protection layer that helps users understand whether a website, wallet address, token, smart contract, or transaction is risky before they connect a wallet or send crypto.
Today, many crypto users only see a wallet address, a token name, or a “Connect Wallet” prompt. In reality, they often do not understand what can happen after they approve the interaction: who they are giving permissions to, where their funds may move next, whether the contract is suspicious, whether the offer is a fake bonus, whether the website is a scam, or whether it could be part of a wallet-drain attack. This is the problem ProofGuard AI is designed to address.
The prototype shows how a modern crypto security application could work: it displays the risk before the user takes action, explains the main warning signals, and recommends whether the user should proceed, verify more information, or reject the interaction completely.
The prototype includes:
a scanner for suspicious crypto websites, wallet addresses, tokens, and smart contracts
dynamic risk evaluation based on the user’s input
a clear risk score and recommended action
a security report with evidence signals and an explanation of the result
demo Google sign-in
a user profile with security history
a protected wallet preview
the “Safe Wallet Guard” concept, designed to show where crypto may move before the user confirms a transaction
The core idea is simple: users should not approve crypto transactions blindly. ProofGuard AI is designed to help surface risk, context, and a possible flow of funds before the user clicks confirm.
This prototype was built directly in Figma Make as a functional competition demo. It uses local demo logic to simulate the behavior of a real SaaS product without connecting to a production blockchain backend. The goal is not only to present a polished landing page, but to show the direction of a product that could help protect users from crypto scams, fake investment offers, scam tokens, suspicious websites, and dangerous wallet permissions.
https://money-last-56707972.figma.site/
2
1
315
TrustFlow — AI Safety & Escrow Layer for Online Marketplaces
1
33
TrustFlow — AI Safety & Escrow Layer for Online Marketplaces
Online marketplaces are powerful, but trust often breaks at the exact moment it matters most: when money, identity, communication, and delivery all meet in one transaction.
TrustFlow is an AI-powered safety and escrow layer designed to protect buyers, sellers, and marketplace platforms before fraud happens.
The product combines:
• AI listing risk analysis
• seller verification and reputation signals
• protected escrow checkout
• scam-warning chat detection
• automatic dispute support
• admin review queues for suspicious cases
I used Figma Make to rapidly move from a rough idea into a published product prototype. The first generated version felt too generic, so I iterated on the visual direction and product story until TrustFlow started to feel like a premium security product: deep-space gradients, subtle star-field atmosphere, glassmorphism cards, AI risk scoring, escrow timelines, warning states, and trust-focused dashboards.
The goal was not just to create a landing page, but to show how AI-assisted design can accelerate real product thinking:
problem → product concept → trust layer → visual identity → published prototype
TrustFlow is built around one simple idea:
Every transaction should feel protected before something goes wrong.
Published prototype:
https://radio-ranch-95224118.figma.site
2
3
700
RoboGarage × TechOps Cockpit
RoboGarage is an AI assistant for automotive workshops, robotics, and vehicle inspection.
The concept connects the full diagnostic workflow: OBD data → DTC codes → live vehicle data → probable cause analysis → safe repair recommendation → service report.
I wanted to create a product system that feels practical, technical, and realistic — not just a visual concept. RoboGarage combines AI-powered diagnostics, robotic inspection, mobile diagnostics, workshop automation, and a professional TechOps Cockpit for managing vehicles, machines, alerts, reports, and repair workflows.
The goal is to show how AI can help workshops diagnose faster, repair more safely, and make better technical decisions while keeping the mechanic in control.
0
141
Designed and developed a modular creator ecosystem with multiple custom editors including story editor, video editor, funnel builder, digital product tools, audio editor, and creator studio workflows.
Built scalable social feed architecture connecting creator content, digital products, and marketplace interactions inside a unified SaaS platform.
0
66
Designed and developed a scalable social feed system for digital products and creator content inside a larger SaaS ecosystem.
Built modular UI architecture for content discovery, creator dashboards, digital product presentation, and responsive feed interactions inspired by modern social platforms.
0
60
Built a real-time messaging platform with modern chat infrastructure, responsive UI, conversation management, live updates, and scalable frontend/backend architecture.
Focused on communication flow, performance, and user experience inspired by modern social platforms.
0
48
Currently building my own automotive diagnostic platform 🚗⚙️
Alongside developing the Trezzo ecosystem, I’ve also been working on a custom automotive diagnostics platform focused on real-time vehicle monitoring and diagnostic workflows.
The system is designed around:
• ECU communication
• Live vehicle telemetry
• Diagnostic trouble codes (DTC)
• System scanning
• Actuator testing
• Coding & service functions
• Hardware integration workflows
The goal is not just UI design, but building a scalable diagnostic environment that combines software architecture, realtime data processing, and automotive tooling into one ecosystem.
Still a work in progress, but it’s exciting to see the platform gradually evolve into a real diagnostic system.
Curious what people from automotive software or embedded systems think about this direction. 👨💻
#Automotive (https://www.linkedin.com/search/results/all/?keywords=%23automotive&origin=HASH_TAG_FROM_FEED) #Diagnostics (https://www.linkedin.com/search/results/all/?keywords=%23diagnostics&origin=HASH_TAG_FROM_FEED) #FullStack (https://www.linkedin.com/search/results/all/?keywords=%23fullstack&origin=HASH_TAG_FROM_FEED) #SoftwareEngineering (https://www.linkedin.com/search/results/all/?keywords=%23softwareengineering&origin=HASH_TAG_FROM_FEED) #RealtimeSystems (https://www.linkedin.com/search/results/all/?keywords=%23realtimesystems&origin=HASH_TAG_FROM_FEED) #VehicleDiagnostics (https://www.linkedin.com/search/results/all/?keywords=%23vehiclediagnostics&origin=HASH_TAG_FROM_FEED) #ECU (https://www.linkedin.com/search/results/all/?keywords=%23ecu&origin=HASH_TAG_FROM_FEED) #Developer (https://www.linkedin.com/search/results/all/?keywords=%23developer&origin=HASH_TAG_FROM_FEED) #UIUX (https://www.linkedin.com/search/results/all/?keywords=%23uiux&origin=HASH_TAG_FROM_FEED) #Bosch (https://www.linkedin.com/search/results/all/?keywords=%23bosch&origin=HASH_TAG_FROM_FEED) #AutomotiveSoftware (https://www.linkedin.com/search/results/all/?keywords=%23automotivesoftware&origin=HASH_TAG_FROM_FEED)
0
48
Building Trezzo Marketplace 🚀
Trezzo is not a template-based project or a quick Playwright-generated demo.
It’s a growing ecosystem built from the ground up:
Marketplace, creator profiles, social feed, digital products, work layer, messaging, payments, affiliate systems, and more — all connected into one platform.
This screenshot shows a real part of the development workflow:
on the left, the actual Vue/Vite source code;
on the right, the live Marketplace running on trezzo.cz (http://trezzo.cz).
The goal is not just to create another website, but to build a scalable ecosystem where every layer works together as a unified product.
#Trezzo (https://www.linkedin.com/search/results/all/?keywords=%23trezzo&origin=HASH_TAG_FROM_FEED) #Marketplace (https://www.linkedin.com/search/results/all/?keywords=%23marketplace&origin=HASH_TAG_FROM_FEED) #VueJS (https://www.linkedin.com/search/results/all/?keywords=%23vuejs&origin=HASH_TAG_FROM_FEED) #Vite (https://www.linkedin.com/search/results/all/?keywords=%23vite&origin=HASH_TAG_FROM_FEED) #SaaS (https://www.linkedin.com/search/results/all/?keywords=%23saas&origin=HASH_TAG_FROM_FEED) #Startup (https://www.linkedin.com/search/results/all/?keywords=%23startup&origin=HASH_TAG_FROM_FEED) #WebDevelopment (https://www.linkedin.com/search/results/all/?keywords=%23webdevelopment&origin=HASH_TAG_FROM_FEED) #Frontend (https://www.linkedin.com/search/results/all/?keywords=%23frontend&origin=HASH_TAG_FROM_FEED) #Ecosystem (https://www.linkedin.com/search/results/all/?keywords=%23ecosystem&origin=HASH_TAG_FROM_FEED) #ProductDevelopment (https://www.linkedin.com/search/results/all/?keywords=%23productdevelopment&origin=HASH_TAG_FROM_FEED)
0
46
Trezzo — Full-Stack SaaS Platform
0
2