Meta

Top Builders

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

General
Company	Meta Platforms, Inc.
Founded	January 4, 2004
Headquarters	Menlo Park, California, U.S.
Repository	https://github.com/facebook

Key Products and Research

Meta has developed a range of AI products designed to enhance various aspects of technology and user experience. Here’s a brief overview of these AI products:

LLaMA (Large Language Model Meta AI)

LLaMA is a series of large language models designed for natural language processing tasks. These models, including the latest LLaMA 3.1, are known for their advanced capabilities in text generation, understanding, and multilingual processing. They are available as open-source models, promoting innovation and research in AI Meta | Social Metaverse Company,Facebook.

Meta AI

Meta AI is an intelligent assistant integrated across Meta’s platforms, such as Facebook, Instagram, WhatsApp, and Messenger. Powered by LLaMA models, it helps users with tasks like content creation, information retrieval, and personalized interactions Meta | Social Metaverse Company.

PyTorch

PyTorch is an open-source machine learning library developed by Meta and widely used in both research and industry. It provides tools for building and training deep learning models and has become a standard framework in the AI community Facebook.

Meta AI Research (FAIR)

Meta’s AI research division, formerly known as FAIR (Facebook AI Research), focuses on advancing the field of AI through open research and collaboration. This division works on various AI challenges, including computer vision, natural language processing, and generative AI Facebook.

Meta AI in the Metaverse

Meta is also incorporating AI into its metaverse initiatives, using AI to create immersive experiences in virtual and augmented reality. This includes developing AI-driven avatars, enhancing virtual environments, and improving interaction within the metaverse Meta | Social Metaverse Company.

AI for Ads

Meta leverages AI to optimize ad targeting, delivery, and measurement across its platforms. AI algorithms analyze vast amounts of data to improve the effectiveness of advertising campaigns, making them more relevant to users and efficient for advertisers Meta | Social Metaverse Company.

LLaMA Impact Grants

The LLaMA Impact Grants program, launched by Meta, aims to support and encourage the innovative use of its LLaMA (Large Language Model Meta AI) models to address critical challenges in various sectors, including education, environmental sustainability, and public good. This initiative offers financial grants and resources to researchers, nonprofits, and other organizations that seek to leverage LLaMA models for impactful projects. The program highlights Meta’s commitment to responsible AI development and its belief in the potential of AI to drive positive social change.

For more details, visit the LLaMA Impact Grants page.

Edit on GitHub

Meta AI Technologies Hackathon projects

Discover innovative solutions crafted with Meta AI Technologies, developed by our community members during our engaging hackathons.

Trade Grid Ai

TradeGrid AI is an autonomous multi-agent trade intelligence platform that revolutionizes how businesses discover, evaluate, and execute cross-border trade opportunities. It replaces static marketplaces and manual research with specialized AI agents that work together to simulate a comprehensive trade intelligence organization. Central to its operation is Band, which allows structured communication and task delegation among the agents. Each agent has a specific role in processing and transforming trade data into actionable insights. The platform utilizes AI/ML APIs for classification, risk detection, and opportunity prediction, enabling intelligent evaluation of suppliers and market demand. Featherless AI serves as the reasoning layer, synthesizing agent outputs into clear trade strategies and market entry recommendations in natural language. The structured workflow includes: 1. A Market Discovery Agent identifies emerging trade opportunities. 2. A Supplier Intelligence Agent organizes and estimates supplier capacity. 3. A Verification Agent assesses trustworthiness through AI-driven scoring. 4. A Matchmaking Agent aligns buyers and suppliers. 5. A Strategy Agent offers actionable trade recommendations. Unlike traditional platforms, TradeGrid AI actively discovers and recommends trade actions in real time, addressing challenges in emerging markets. It facilitates data-driven decisions across sectors like logistics, procurement, agriculture, and manufacturing. By integrating collaborative agent networks, TradeGrid AI transforms fragmented trade research into a cohesive ecosystem that continuously generates actionable business insights from global market data.

PulseIntel — Enterprise Web Intelligence Platform

PulseIntel is a dual-track enterprise web intelligence platform built for modern security and revenue teams. Every company has two critical blind spots. First, competitors are quietly hiring machine learning engineers and fraud analysts — revealing their next product move months before any announcement. Nobody has time to read thousands of job postings and connect the dots manually. Second, phishing pages, fake domains, and credential dumps appear on the open web daily. Security teams find out only after customers start complaining because internal tools cannot monitor what lives outside the firewall. PulseIntel solves both problems with one unified platform. Track 1 — GTM Intelligence: Enter any company name and PulseIntel uses Bright Data MCP Server to scrape job postings across LinkedIn, Greenhouse, Lever, and company career pages in real time. Groq AI running LLaMA 3.3 70B analyzes the hiring patterns and generates a structured competitive strategy brief — what the company is building, which departments are growing, which competitors should be concerned, and an expected timeline. Track 3 — Security and Compliance: Enter any brand name and PulseIntel scans paste sites, social media, and the open web for brand mentions using Bright Data MCP Server. Groq AI scores each finding by risk level from 1 to 10 and categorizes threats as phishing, credential leak, lookalike domain, or impersonation. Each alert includes a recommended action for the security team. The entire data pipeline runs on Bright Data MCP Server which bypasses bot detection, handles JavaScript rendering, and returns clean markdown directly consumable by the AI layer. The dashboard is built on Streamlit with a dark enterprise aesthetic, real-time metrics, and risk-level filtering. PulseIntel was built in 4 days by WeCoders for the Web Data UNLOCKED Hackathon 2026.

Sentinel Web-Risk Intelligence Platform

Modern enterprises rely on thousands of global vendors, yet existing risk tools depend on stale audits and self-reported data. By the time a vendor failure is detected, catastrophic operational damage has already occurred. Sentinel Web-Risk Intelligence Platform solves this with a Zero-Trust Autonomous Investigation Architecture driven by real-time voice telemetry. Instead of manual lookups, analysts trigger end-to-end deep dives hands-free. Using a client-side Speechmatics streaming engine, the platform instantly ingests, scrubs, and translates conversational audio commands into target queries with sub-100ms latency. Upon voice activation, Sentinel unleashes six specialized CrewAI agents that autonomously cross-examine the live open web using Bright Data infrastructure. First, the Recon Agent fires targeted search queries through the Bright Data SERP API. Next, the Scraping Agent extracts data from JS-heavy portals and CAPTCHA-shielded filings using the Scraping Browser and Web Unlocker, routing traffic dynamically through Bright Data's premium Global Proxy Network to bypass geo-blocks and scan restricted regional registries. Then, the Verification Agent filters out noise by assigning strict credibility scores to every harvested text source. Following this, the Intelligence Agent synthesizes verified signals across five critical corporate dimensions: Financial, Operational, Legal, Reputational, and Cybersecurity. The Prediction Agent then models an accurate 90-day disruption probability timeline using machine learning patterns. Finally, the Reporting Agent compiles these analytics into a structured, C-suite ready intelligence dossier. The final output delivers a dynamic 0–100 risk score, predictive forecasting models, and clear remediation paths. This telemetry streams instantly through an automated WebSocket dashboard by flawlessly unifying Speechmatics voice routing with Bright Data's unblockable scraping pipelines.

SalesPilot AI —The Autonomous B2B Sales Researcher

SalesPilot AI: Autonomous B2B Sales Research The Problem B2B Account Executives and SDRs spend up to 30 percent of their day manually researching prospects across LinkedIn, company websites, and news just to see if an account is worth pursuing. This manual process is slow, error-prone, and doesn't scale. The Solution SalesPilot AI automates the pre-sales research workflow. Built with a sophisticated multi-agent architecture, it operates like a team of researchers. When a user inputs a target company, AI agents dynamically scrape the web, extract data, and synthesize unstructured information into a cohesive intelligence report. Key Features: Deep Company Profiling: Extracts leadership details, business models, and industry positioning automatically. Buying Intent Scoring: Analyzes hiring velocity, recent funding rounds, and pain points to calculate a Buying Intent Score, helping reps prioritize accounts actively looking for solutions. Tech Stack & Competitor Intel: Identifies the software a company uses and pinpoints competitor displacement opportunities. Company Comparisons: Run head-to-head comparisons between two companies to identify the better prospect based on specific dimensional breakdowns. Automated Outreach Generation: Drafts hyper-personalized outreach emails tailored specifically to the company's pain points and recent news. CRM Integration: Syncs the enriched data directly to HubSpot with a single click. How it was built The backend is a high-performance Python application powered by FastAPI, LangGraph, and a massive LLM waterfall system routing between NVIDIA NIM, Google Gemini, Groq, Cerebras, and OpenRouter. The intelligence agents utilize Bright Data's Web Unlocker and SERP APIs to bypass anti-bot protections and gather real-time data. The frontend is a beautifully designed Next.js application, featuring real-time Server-Sent Events that stream the agents' live progress to the user so they can watch the AI research in real time.

SignalOS - AI-Powered GTM Intelligence Platform

SignalOS is an AI-powered GTM (Go-To-Market) Intelligence Platform designed to help sales, marketing, and revenue teams make faster and smarter decisions using real-time market intelligence. Today, GTM teams spend countless hours manually tracking competitors, monitoring hiring activity, analyzing product launches, researching market trends, and identifying buying signals across multiple websites and platforms. SignalOS automates this entire workflow. The platform continuously gathers intelligence from web sources, company updates, hiring trends, product announcements, social discussions, and market signals. Using an AI-powered synthesis engine, SignalOS transforms raw data into structured intelligence reports containing executive summaries, competitor analysis, product activity, hiring insights, risks, opportunities, and buying intent indicators. Users can track companies, run intelligence scans, monitor market movements, and receive actionable recommendations without performing manual research. SignalOS also provides a centralized dashboard for visualizing signals, competitor activity, and intelligence history. Our goal is to replace fragmented market research workflows with a single intelligent platform that converts publicly available information into revenue-driving insights. By helping teams identify opportunities earlier, react to competitor moves faster, and prioritize high-intent accounts, SignalOS enables organizations to operate with greater speed, confidence, and efficiency in an increasingly competitive market.

AEGIS Sentinel: Multi-Agent Emergency Terminal

AEGIS Sentinel is a premium, real-time crisis coordination and spatial tracking dashboard built to solve communication blindness during regional disasters. During natural emergencies, survivors and dispatchers face network outages, rumor pollution, and static, cross-country evacuation paths that lead directly into hazard zones. AEGIS Sentinel addresses these challenges by deploying a collaborative, parallel multi-agent pipeline. The ASR Ingestion Agent processes dispatcher feeds using Speechmatics' Enhanced ASR. The Web Search Agent scrapes live reports matching the geolocalized city using Bright Data and Google News RSS feeds. The AI Reasoning Agent (Gemini 3.5) filters out rumors, evaluates threat details, extracts coordinates, and assigns verification scores. Finally, the Automation Agent registers threats in Cognee Graph Memory and triggers custom webhooks. Key Features: - One-Click Geocentric Localization: Automatically centers Leaflet dark maps via network/GPS/IP Geolocation, reverse-geocodes via OSM Nominatim, and queries custom Indian Cities and GeoDB API caching routes. - Proximity-Constrained Evacuation: If standard shelters are over 50km away, the system projects a virtual municipal shelter 1.5km away, drawing neon-green dashed paths. - Acoustic SOS Siren: Employs Web Audio API oscillators to emit European safety horns for search-and-rescue. - Walkie-Talkie Speech warnings: Synthesizes automated alerts with simulated radio beeps and trailing white-noise static squelches. - Live Background SMTP Dispatch: Fires comprehensive advisory emails listing active threats directly to operator inboxes asynchronously.