Replit

Top Builders

Explore the top contributors showcasing the highest number of app submissions within our community.

Replit

Replit is an integrated development environment (IDE) and collaboration platform designed to make coding more accessible and seamless for developers of all levels. It allows users to write, execute, and share code in over 50 languages directly from a browser. The platform stands out for its ease of use, allowing developers to instantly start coding without complex setup procedures. Replit also supports collaborative coding, version control, and deployment, making it ideal for both individual developers and teams.

General
Author	Amjad Masad, Replit Founder & CEO
Release Date	Initial release in 2016
Website	https://replit.com
Repository	https://github.com/replit
Documentation	https://docs.replit.com/
Technology Type	Cloud-based IDE and collaborative development platform

Key Features

Multilingual Support: Replit offers support for 50+ programming languages, including Python, JavaScript, Ruby, and C++, making it a versatile platform for diverse development needs.
Collaborative Coding: Replit provides real-time collaboration tools, enabling multiple users to code together in a shared environment, much like Google Docs for coding.
Integrated GitHub Support: Users can import and manage GitHub repositories directly within Replit. This allows seamless version control and easy collaboration across platforms.
Replit AI: An AI-powered assistant that helps developers by generating code, explaining errors, and providing suggestions to improve productivity.
Deployments: The platform allows developers to deploy full-stack applications directly from Replit, simplifying the process of launching web applications.
Replit Bounties: A marketplace where developers can post or take on freelance coding tasks, making it a useful feature for those looking to earn through coding or get assistance with projects - Replit Docs.

Replit AI Agent

Replit AI Agent is a powerful tool designed to enhance development workflows by acting as an AI assistant integrated within the Replit platform. It enables developers to automate tasks, streamline coding processes, and interact with projects using natural language commands. With the ability to run, test, and debug code directly within Replit, the AI Agent can also write code snippets, make suggestions, and even collaborate in real-time, making the development experience faster and more intuitive.

👉 For more details, you can explore Replit's documentation here.

Start Working with Replit

Getting started with Replit is simple:

Sign Up: Visit Replit and sign up for an account. You can start coding right away, as no additional installation is required.
Create or Import Projects: You can either create a new project ("Repl") or import a project directly from GitHub by connecting your account and selecting the repository to clone. Detailed instructions on using Git with Replit can be found here.
Start Coding: Choose your preferred programming language and dive into coding. Replit's real-time collaborative features allow you to invite team members to work together.
Deploy Your Project: Once you're ready, Replit allows easy deployment of web apps or services with just a few clicks.
Explore Replit Bounties: You can also earn or hire talent by using Replit Bounties.

Edit on GitHub

Replit AI technology page Hackathon projects

Discover innovative solutions crafted with Replit AI technology page, developed by our community members during our engaging hackathons.

UvicornForge-AI

UvicornForge AI is an AI-powered co-founder designed for hackathon teams. Enter a project idea along with real parameters such as team size, total funding, available time, target users, industry, and technologies (including AMD GPUs and Fireworks AI). The system generates a comprehensive, structured startup brief covering problem, solution, MVP scope, key features, demo scenario, business model, risks, go-to-market strategy, and more. It also produces ready-to-use artifacts including a pitch deck outline, full demo script, MVP checklist, and starter README. A key innovation is the Success Score (1-10) predicted by a custom MLP model. Unlike generic scores, ours is trained on a high-signal target engineered to respond to actual project parameters: larger teams, realistic funding, longer development time, and explicit AMD usage all meaningfully increase the predicted success. The frontend allows direct input of these values, and the feature mapper feeds them straight into the model along with an ambition factor extracted from the idea description. The backend combines Fireworks AI for high-quality text generation with a locally running PyTorch MLP (65 features) trained on an AMD-tailored dataset of 10,000 examples. We achieve strong validation performance (R² ≈ 0.85) while ensuring the score differentiates between weak and strong proposals. The modern frontend features live model metrics, smooth animations, and one-click downloads of all generated materials. Built specifically for the Unicorn Track, the project demonstrates practical use of AMD GPUs for ML inference and Fireworks AI for LLM generation in a real, useful product that helps teams create better submissions faster.

Curio

Paste this at the end of the text you already have: Development Philosophy During the development of Curio, our primary focus was on "Time-to-Value." We understood that in a fast-paced hackathon environment, the ability to iterate quickly and show tangible results is paramount. By leveraging the latest Google GenAI SDK, we were able to abstract away complex model interactions and focus on creating a user experience that feels premium and responsive. The decision to integrate marked.js was born out of a necessity to present AI-generated content in a human-readable format, bridging the gap between raw data and actionable information. Our development process involved rigorous testing of the frontend-backend communication, ensuring that even under high load or potential API timeout scenarios, the application remains stable and visually consistent. Future Roadmap While Curio is already a functional intelligence platform, we envision several paths for future development to enhance its impact: Multi-Modal Integration: Expanding beyond text to provide video summaries and audio-briefing capabilities for users on the go. Granular Preference Learning: Implementing an automated feedback loop where the AI learns from the articles a user clicks on, refining its filtering logic over time to become increasingly accurate. Enterprise Customization: Creating a B2B version of Curio that allows companies to ingest internal company news and market reports, providing employees with a personalized feed of internal announcements alongside industry trends. Global Language Support: Expanding the linguistic capabilities of the Gemini model integration to serve non-English speaking markets, ensuring that personalized news is accessible regardless of the user's primary language.

SMC Liquidity Hunter

SMC Liquidity Hunter is the institutional-grade decision layer retail traders and AI trading agents don't have. Retail platforms — including Robinhood's own Agentic Trading, launched in May 2026 — are giving AI agents real trading power, but those agents have no visibility into the market structure institutions actually trade on. SMC Liquidity Hunter closes that gap. A real-time Smart Money Concepts engine continuously computes order blocks, fair value gaps, liquidity sweeps, and break-of-structure as live crypto and forex data streams in. That analysis is exposed as callable tools through a genuine MCP server — verified end-to-end with a real JSON-RPC handshake against our live deployment — so any external AI agent can query institutional-grade market reasoning on demand. The reasoning layer runs on Gemma 4 (26B A4B, MoE) served via vLLM on AMD ROCm, on a live MI300X instance on AMD Developer Cloud. We chose the MoE variant specifically for its low-latency tool-calling performance in the agent loop. This isn't a concept — it's proven. We placed a real Alpaca paper order end-to-end: order submitted, order_id returned, status read back as FILLED at live market price, and a cancel-after-fill attempt correctly rejected by the broker. Every core claim in this submission — the MCP handshake, the AMD deployment, the broker execution — was independently verified against live, running infrastructure, not just described.

kuber ai

Kuber AI Transforming Personal Finance Through Enterprise-Grade Intelligence In an era where financial complexity is at an all-time high, the average individual often struggles to maintain a cohesive view of their economic health. While enterprise-level organizations utilize sophisticated, integrated systems to monitor cash flow, asset allocation, and long-term liabilities, individuals are frequently left to juggle fragmented spreadsheets, disjointed banking apps, and disconnected investment portfolios. Kuber AI was conceived to bridge this gap, bringing the power of institutional-grade financial oversight directly to the fingertips of the individual user.

Compliance Guardian

`Compliance Guardian is an automated multi-agent financial compliance and transaction monitoring system designed to address the high costs and inefficiencies of BSA/AML compliance. Built for the Band of Agents Hackathon 2026, the project demonstrates how autonomous AI agents can collaborate securely to run high-stakes regulatory workflows with a built-in human-in-the-loop validation layer. The system orchestrates three specialized agents communicating through the Band coordination layer: Risk Analyst Agent: Automatically scans incoming transactions, calculates a risk score based on rule sets, and flags anomalies. Compliance Officer Agent: Evaluates flagged transactions, detects SAR (Suspicious Activity Report) filing requirements, and drafts compliant case details. Human Review Agent: Escalates high-risk cases to a real-time web dashboard for manual approval or rejection. By separating duties across collaborative agents and logging every message exchange, the system maintains a complete audit trail of decisions. This approach drastically reduces the time needed to review alerts, minimizes human error, and ensures complete compliance ready for regulatory audits.`

CareBand — 5-Agent Healthcare Coordination System

CareBand is a 5-agent healthcare coordination system built for Track 3: Regulated & High-Stakes Workflows. It addresses a critical gap in chronic patient care: 1.4 billion patients worldwide, 50% miss medications regularly, and families find out 4–6 hours too late when emergencies happen. The system deploys five specialized AI agents that collaborate exclusively through a central message bus. Triage detects symptoms and assigns risk levels. Medication checks drug interactions and missed-dose severity. Guardian makes escalation decisions (MONITOR, ESCALATE_SCHEDULED, or ESCALATE_IMMEDIATE). Family alerts caregivers in real time via SMS, WhatsApp, or in-app notifications. Compliance logs every decision with immutable audit trails for GDPR and HIPAA. Every agent communicates through structured JSON handoffs — not chat, but real task delegation with context sharing. Band serves as the core collaboration layer: Triage sends to Band, Medication reads from Band, Guardian reads from Band, Family reads from Band, and Compliance reads all messages from Band. Remove Band, and the system collapses. The architecture is fully implemented with real-time SSE streaming, fire-and-forget message routing, and full pipeline execution. The current prototype uses a local message store that mirrors Band's exact routing logic, with live Band API integration pending OAuth token configuration — a transport-layer switch requiring zero code changes.