DATE

2025-2026

DATE

2025-2026

DATE

2025-2026

CoSparq

An AI platform that evaluates and develops startup ideas, delivering structured feedback and AI chatbot for idea development.

SaaS

SaaS

SaaS

AI Powered

AI Powered

AI Powered

Type

SaaS full stack application

Primary Technologies Used

JavaScript, Node.js, Express.js, HTML, CSS, OpenAI API, ElevenLabs, Google Oauth2.0

Owners

Marwan Youssef Abdelrazik

Ultra Watch
Ultra Watch
Ultra Watch
Ultra Watch

What is CoSparq?

CoSparq is an AI-powered platform designed to help founders validate, build, and launch ideas faster. It combines agentic AI systems, real-world automation, and full-stack engineering to turn raw ideas into production-ready products—without the traditional friction of early-stage development.

Agentic Systems

Engineered a context-aware Agentic AI chatbot and voice Agent with persistent memory layered on top of stateless LLM APIs. The AI maintains long-term conversation context, user intent, and project state across sessions. Supports complex, multi-step reasoning instead of single-turn prompt responses.

The Purpose of CoSparq

Most founders waste time stitching together tools, validating ideas too late, or rebuilding the same infrastructure repeatedly. CoSparq centralizes ideation, validation, execution, and automation into a single AI-driven workflow—allowing founders to move faster with fewer resources and better technical foundations.

Ultra Watch
Ultra Watch
Ultra Watch
Ultra Watch
Ultra Watch Zoom Shot
Ultra Watch Zoom Shot
Ultra Watch Zoom Shot
Ultra Watch Zoom Shot

Technical: The Technologies I Used and What I Learnt

Building CoSparq required designing scalable AI systems, secure infrastructure, and cost-efficient workflows while integrating multiple external APIs. This project pushed me beyond simple LLM usage into full agentic system design and production engineering.

AI & External Service Integration

— Integrated Twilio and ElevenLabs for real-world voice interactions
— Used TwelveLabs with Gemini AI for video analysis and safety assessment
— Connected AI outputs into user-facing tools and conversational interfaces

The site was built using modern web technologies to ensure performance and responsiveness. I implemented a user-friendly content management system (CMS) that allows the studio to easily update services, showcase new collections, and manage client bookings. The site is fully responsive, providing a seamless experience across all devices, from desktop to mobile.

The site was built using modern web technologies to ensure performance and responsiveness. I implemented a user-friendly content management system (CMS) that allows the studio to easily update services, showcase new collections, and manage client bookings. The site is fully responsive, providing a seamless experience across all devices, from desktop to mobile.

The site was built using modern web technologies to ensure performance and responsiveness. I implemented a user-friendly content management system (CMS) that allows the studio to easily update services, showcase new collections, and manage client bookings. The site is fully responsive, providing a seamless experience across all devices, from desktop to mobile.

Backend & AI Architecture

  1. Implemented Stripe subscriptions with tiered plans and secure webhook verification.

  2. Built per-user API cost tracking across 15+ OpenAI models in real time.

  3. Designed the system with security-first principles: JWT authentication, HTTP-only cookies, refresh token rotation, role-based access control, and API rate limiting.