AI SaaS Builder. Agents, RAG & Custom AI Products creator
New to Contra
AI SaaS Builder. Agents, RAG & Custom AI Products creator
AI Tender Evaluation Platform - Multi-Agent Scoring
Designed and built an AI-powered procurement platform for automated tender generation, compliance validation, and vendor ranking. The system uses multi-agent orchestration to extract structured requirements, validate vendor responses via RAG, and compute weighted evaluation scores based on government scoring standards. Includes configurable scoring methods (relative, 0–10 scale, government models), detailed comparison matrix, and AI-assisted requirement drafting. Built as scalable SaaS for enterprise and government procurement workflows.
0
40
Hi Contra, I’m Ivan, founder of unrealos, an AI product development studio.
We build custom AI SaaS platforms, AI agents, RAG systems, and business automation tools for companies across the US, UK, Australia, Canada, and Asia.
Over the past few years, we have delivered 30+ AI products, from early MVPs to complex multi-tenant platforms used in real business operations.
Here I’ll share our projects, practical lessons from building AI products, and what actually matters beyond connecting another LLM API.
Glad to join the community.""
1
52
Energy AI Workforce & Qualification Management Platform
Built an AI-powered workforce and document management platform for an offshore energy company operating across multiple projects and locations.
The platform centralizes employee CVs, technical qualifications, project assignments, and certification records. It helps operations teams quickly find suitable technicians, track qualification validity, manage renewals, and access workforce documents without searching through folders and spreadsheets.
Key features include:
- AI-assisted CV and document search
- Technician filtering by skills and qualifications
- Qualification and certification tracking
- Renewal workflows with document re-upload
- Personnel allocation across active projects
- Employee status management, including suspended workers
- Project lifecycle and workforce visibility
- Automated email notifications
- Role-based access for operations teams and supervisors
- Centralized cloud infrastructure on Google Cloud and Firebase
The solution reduced manual administration, improved visibility across the workforce, and created a scalable foundation for company-wide rollout.
My role: Product architecture, AI solution design, technical leadership, and delivery management.
Tech stack: Firebase, Google Cloud, Claude API, JavaScript, HTML/CSS, Cloud Functions, Google Workspace integrations.
0
45
AI Deep Company Research SaaS
Built a custom AI-powered company research platform for a B2B sales consultant to automate deep prospect research before client meetings and sales calls.
The system collects and analyzes information from company websites, LinkedIn, Crunchbase, industry sources, and recent news. It then converts the raw data into a structured strategic report covering the company’s business model, market position, key initiatives, potential challenges, relevant decision-makers, and possible sales opportunities.
The platform includes:
- Automated multi-source company research
- AI-generated strategic analysis and opportunity mapping
- Structured, repeatable research workflows
- Custom report generation with downloadable PDF output
- Research history and report management
- Source validation and structured data extraction
- Scalable backend architecture for processing long research tasks
The main goal was not simply to summarize public information, but to help the client understand how to approach each prospect, what business outcomes to discuss, and where a potential engagement could create value.
As a result, the platform reduced average pre-call preparation time from around 2 hours to approximately 20 minutes, while making the research process more consistent across different companies and industries.
Tech stack: Python, FastAPI, React, Supabase, PostgreSQL, Perplexity AI, GPT, background processing, PDF generation.
Welcome to Mirimera!
Mirimera Official Website.
Built with our own react component library.
Tech Stack:
- Next JS
- React Aria
- Tailwind CSS
- TypeScript
Another day, another step towards becoming a better design engineer.
https://cryptodashboardds5.netlify.app/
1
1
100
Just reimagined Grok's landing page with the help of AI and magic path.
https://grokreimagined.netlify.app/
2
108
Meet Sticker Studio, drop in any SVG and it turns into a glossy die-cut sticker you can actually peel. 🎨
The fun part: I built the whole thing in Wonder. Designed the UI in Wonder Chat, generated the outline shaders in Wonder, then used Wonder's MCP to pipe it all into Claude Code as a real Next.js app. The shader-filled outlines you see? Those are live Wonder shaders masked to the die-cut edge.
Even this video started in Wonder, I mood boarded and story boarded the whole thing in Wonder first, then built it as a Remotion composition that renders the app's actual stickers.
🎨 Wonder file: Wonder file link
(https://app.wonder.so/muhammad-ammar/files/019f173a-c717-700d-a158-fa34a4b44f0e)🔗 Try it: stickinglab.netlify.app
(http://stickinglab.netlify.app)🎥 Walkthrough: Loom link
(https://www.loom.com/share/da8c20ae9d9445d1baaac82b08b24e78)
2
311
I'm building an icon library that refuses to sit still
I have a soft spot for the details most people scroll right past. The hover state. The little bounce when something lands. The blink you only notice the second time.
So I started building an entire icon library around that idea, and I wanted to share an early look while it is still taking shape.
Around 30 icons in so far, each one animated with real spring motion it's still WIP