Freelancers using AgentKit in India
Freelancers using AgentKit in India
Sign Up
Post a job
Sign Up
Log In
Filters
2
Projects
People
Subhradip Roy
pro
Bengaluru, India
AI Web App & SaaS Developer | Agents and AI UGC Creator
$1k+
Earned
3x
Hired
4.9
Rating
24
Followers
Follow
Message
AI Web App & SaaS Developer | Agents and AI UGC Creator
0
ClearBook - AI SaaS Development | For Enterprise Data Management
0
19
0
AI Powered Fitness Web App | $500+ / yr per User Cost Saving
0
19
3
AI-Powered Virtual Try-On for Multi-Brand E-Commerce ● Our platform explores a different approach: a virtual twin generated from the user’s profile and photo, allowing customers to try on pieces instantly and build complete outfits directly on a lifelike preview. Instead of imagining how items might work together, shoppers can see the full look in motion before committing. ● The interface prioritizes construction over browsing. Products can be applied to the twin with a single action, while intelligent suggestions help complete the outfit without interrupting the flow. A live “Look Total” keeps track of the combined price, making it easy to move an entire outfit to checkout at once. ● Reducing uncertainty while keeping the experience calm, intuitive, and familiar to traditional e-commerce users.
3
352
0
AI-Powered Real Estate SaaS Platform That Cuts Costs, Closes Deals Faster & Maximizes ROI ● An AI-powered real estate platform that simplifies every step of buying or investing in property. From smart search and mortgage options to valuation, legal handling, and closing, everything is unified in one seamless ecosystem. ● Our AI assistant provides tailored recommendations, real-time market insights, and transparent pricing so users can save money and avoid traditional agent commissions. ● With automated workflows and data-driven guidance, makes real estate faster, clearer, and accessible to everyone - whether you're purchasing a home or exploring investment opportunities.
0
284
AgentKit
(1)
Follow
Message
Kaustub Singh
India
Engineer.
$1k+
Earned
1x
Hired
Expert
Follow
Message
Engineer.
0
Vidlume — AI Video Ads From Photos & Clips What problem does Vidlume solve? Most businesses already have product photos, property images, clips, brochures, and basic listing details — but turning those assets into high-quality short-form videos is slow, expensive, and inconsistent. Vidlume solves the problem of creating polished video ads regularly without needing an editor, scriptwriter, voiceover artist, designer, or social media team. The platform is positioned for businesses that need ready-to-post content for Instagram Reels, YouTube Shorts, TikTok, and WhatsApp. Core problems Vidlume solves 1. Businesses struggle to post consistently Short-form platforms reward frequent posting, but most small teams do not have time to create professional video content every day. Vidlume helps turn existing visuals into repeatable daily content. 2. Video creation is expensive Hiring video editors, scriptwriters, voiceover artists, and social media managers every month can become costly. Vidlume replaces much of that manual workflow with AI-assisted automation. 3. Static photos do not stop the scroll Normal product photos or property images often look plain on social media. Vidlume transforms those assets into videos with motion, hooks, captions, music, voiceover, branding, and CTAs. 4. Real estate listings need better storytelling For real estate brokers, static listing photos are not enough. Listings need room-by-room flow, location context, pricing, amenities, urgency, and a clear call-to-action. Vidlume is built to generate branded property reels, WhatsApp promos, listing showcases, investment reels, and locality explainers. 5. Generic templates make brands look low-effort Instead of simple slideshow templates, Vidlume creates structured video formats such as UGC-style ads, product demos, offer-first videos, luxury property reels, investment-angle reels, and locality-based content. What Vidlume does Vidlume takes the assets a business already has and turns them into complete short-form video ads. Users can upload photos, short clips, product visuals, property images, listing details, price, location, amenities, contact details, logo, and brand information. The platform then generates a ready-to-post video with: AI-written script Human-like voiceover Captions and subtitles Scene sequencing Motion effects Transitions Background music Branding CTA screen Multi-format exports Scheduling and publishing support The website describes Vidlume as a platform that handles scripting, voiceover, captions, music, transitions, formatting, scheduling, and creative variations from existing visuals. How I built Vidlume technically Frontend Built with Next.js, giving users a clean dashboard where they can upload assets, manage projects, preview generated videos, edit scripts, choose formats, and export final content. Backend Built with Node.js, handling project creation, asset processing, AI workflow orchestration, video generation requests, user data, and API communication between the frontend, database, AWS, and AI services. Database Uses PostgreSQL to store users, projects, uploaded assets, generated scripts, video metadata, brand settings, export status, and scheduling data. Cloud infrastructure Uses AWS for scalable media handling, including secure file storage, processing workflows, and job execution. Queue system Uses AWS SQS to manage video generation jobs reliably. This allows heavy processing tasks to run asynchronously without blocking the user interface. AI integrations Multiple AI integrations are used for: Understanding uploaded images and clips Detecting scenes such as kitchen, bedroom, exterior, balcony, product close-up, etc. Generating marketing scripts Creating hooks and CTAs Producing voiceovers Generating captions and overlays Creating different creative variations from the same assets FFmpeg video processing Uses FFmpeg for complex video generation, including: Stitching images and clips together Adding zoom and motion effects Applying transitions Syncing visuals with voiceover Adding background music Burning captions and overlays Adding logos and CTA screens Exporting in 9:16, 1:1, and 16:9 formats Optimizing videos for Reels, Shorts, TikTok, and WhatsApp Simple user-facing pitch Vidlume helps businesses turn everyday photos and clips into professional short-form video ads automatically. Instead of spending hours in Canva, CapCut, or paying editors every month, users can upload their assets once and generate ready-to-post videos with scripts, voiceovers, captions, music, transitions, branding, and CTAs. For real estate, Vidlume turns property photos into premium listing reels with room-by-room flow, location hooks, price overlays, amenities, broker branding, and WhatsApp CTAs. For product businesses, it creates product showcases, UGC-style ads, offer videos, and social-ready creatives. Best presentation version for potential users The Problem Creating good video ads consistently is hard. Most businesses have photos and clips, but they do not have the time, skills, or budget to turn them into professional videos every week. The Solution Vidlume uses AI to convert existing visuals into ready-to-post video ads with scripts, voiceovers, captions, branding, music, transitions, and clear CTAs. Who it is for Vidlume is built for real estate brokers, builders, property marketing agencies, Shopify sellers, D2C brands, creators, social media managers, and local businesses that need consistent video content. Why it matters Businesses can create more content, test more hooks, post more consistently, and make their products or listings look more professional without hiring a full creative team. Technical trust Behind the scenes, Vidlume is powered by Node.js, Next.js, PostgreSQL, AWS, SQS, multiple AI integrations, and FFmpeg-based video processing. This allows the platform to handle everything from asset upload to AI analysis, script generation, voiceover creation, video stitching, branding, export, and scheduling.
0
24
0
Lightweight Online Video Editor A fast, lightweight online video editor for people who need quick media edits without installing heavy editing software. Users can upload a video, make simple edits, preview changes, and export a ready-to-share video directly from the browser. It is designed for creators, marketers, small businesses, social media managers, and anyone who wants quick video editing without complexity. Features It Offers Trim and cut videos Merge multiple clips Resize videos for social platforms Convert videos to different formats Add text and captions Add background music Remove or replace audio Crop and rotate videos Compress videos for faster sharing Generate thumbnails Export videos in high quality Browser-based editing with no installation Technical Details The system is built with a simple and scalable video-processing workflow. Frontend Next.js / React interface Drag-and-drop video upload Timeline-based preview Edit controls for trim, crop, resize, text, audio, and export Responsive UI for desktop and mobile Backend Node.js backend PostgreSQL database AWS S3 for storing uploaded and processed videos Queue system using AWS SQS or BullMQ Background workers for video processing User project and export history management Video Processing FFmpeg for all complex media processing Trimming and cutting clips Merging videos Resizing and cropping Audio mixing Format conversion Video compression Thumbnail generation Caption and text overlay Final video rendering Storage & Export Uploaded videos are stored securely Edited videos are processed in the background Final exports are saved and made available through a shareable download link How It Works User uploads a video User applies quick edits in the browser Edit settings are sent to the backend FFmpeg processes the video based on selected edits Final video is exported and saved User downloads or shares the finished video Business Value This editor helps users: Save time on simple edits Avoid heavy editing software Create social-ready videos faster Reduce editing cost Work directly from the browser Export clean, ready-to-post content Quick video edits in your browser — powered by FFmpeg.
0
14
0
AI Customer Support Agent for Businesses What i Built We built an AI customer support agent that helps businesses answer customer questions instantly, reduce repetitive support tickets, and support customers 24/7. The AI is trained on the business’s own content, including website pages, FAQs, help docs, policies, pricing, product details, and past support conversations. It does not work like a basic chatbot. It works like a trained support assistant that understands customer questions, finds the right information, replies in the brand’s tone, and escalates complex issues to a human. Problem It Solves Businesses spend too much time answering the same customer questions again and again: Where is my order? What is your refund policy? How do I cancel? What is included in this plan? How do I book an appointment? Can I speak to support? This creates slow response times, higher support costs, and frustrated customers. Our AI agent solves this by handling repetitive support automatically. What the AI Does When a customer sends a message, the AI: Understands the customer’s question Detects the intent, such as refund, order, billing, pricing, or technical issue Searches the business knowledge base Finds the most relevant answer Responds in the company’s brand tone Asks follow-up questions if needed Escalates to a human when the issue is complex or sensitive Summarizes the conversation for the support team This helps customers get quick answers while the team focuses on important issues. Key Features AI trained on business data Website chat widget FAQ and document upload Website content crawling Instant customer replies Human handoff Conversation summaries Brand tone customization Intent detection Support analytics Multi-language support Helpdesk and CRM integrations Order, billing, and ticket lookup through APIs Technical Architecture The system uses a retrieval-based AI architecture. When business content is uploaded, the system: Extracts text from websites, PDFs, FAQs, and documents Splits the content into small chunks Converts each chunk into embeddings Stores them in a vector database Searches the most relevant chunks when a customer asks a question Sends the retrieved context to the AI model Generates a safe and accurate response based on approved business knowledge This reduces hallucination and keeps answers business-specific. Recommended Tech Stack Frontend Next.js React Tailwind CSS Embedded JavaScript chat widget Admin dashboard Backend Node.js / Express PostgreSQL Prisma ORM Redis Queue system using BullMQ or AWS SQS AWS S3 for document storage AI Layer LLM for response generation Embedding model for knowledge search RAG pipeline Intent detection Sentiment detection Conversation summarization Escalation logic Vector Database PostgreSQL with pgvector for MVP Pinecone, Qdrant, or Weaviate for scaling Integrations Shopify WooCommerce Stripe Zendesk Freshdesk Intercom HubSpot Slack WhatsApp Business API Human Handoff The AI escalates to a human when: It cannot find a confident answer The customer asks for a human The issue involves refund approval The customer is angry The issue is related to billing, cancellation, or legal concerns Account-specific action is required Before handoff, the AI creates a short summary so the support team can respond faster. Business Value The AI support agent helps businesses: Reduce repetitive tickets Reply to customers instantly Lower support workload Provide 24/7 support Improve customer satisfaction Maintain consistent support quality Discover common customer problems through analytics Final Positioning An AI customer support agent that answers customers instantly, reduces repetitive tickets, and hands off complex issues to your team. Or shorter: Scale customer support without hiring more agents.
0
34
0
AI Sales & Lead Qualification Agent What We Built We built an AI sales agent that helps businesses capture, qualify, and convert leads automatically. Instead of letting website visitors leave without action, the AI instantly starts a conversation, asks the right sales questions, understands buyer intent, collects contact details, scores the lead, and books meetings with the sales team. It works 24/7, so businesses never miss high-intent leads. Problem It Solves Most businesses lose leads because: They reply too late Their team cannot respond 24/7 Unqualified leads waste sales time Hot leads are not followed up properly Website visitors leave without booking a call Our AI agent solves this by responding instantly and qualifying leads before they reach the sales team. What the AI Does When a visitor lands on the website, the AI: Starts a natural sales conversation Understands what the lead is looking for Asks qualifying questions Collects name, email, phone, company, and requirements Detects buying intent and urgency Scores the lead as cold, warm, or hot Recommends the best next step Books a meeting with the sales team Sends the lead details to CRM Creates a short lead summary for follow-up Example Flow Visitor: “I’m interested in automating our sales follow-up.” AI Agent: “Great. How many leads does your team handle every month?” Visitor: “Around 500 leads.” AI Agent: “Got it. Are you currently using any CRM like HubSpot, Salesforce, or Zoho?” Visitor: “Yes, HubSpot.” AI Result: Lead Score: High Intent: Strong Next Step: Book sales call Summary: Company handles 500 monthly leads and wants sales follow-up automation with HubSpot integration. Key Features AI sales conversation Lead qualification Buyer intent detection Lead scoring Meeting booking CRM integration Human handoff Lead summary generation 24/7 website response Follow-up message suggestions Sales dashboard Conversation history Technical Details The system is built using an AI agent workflow. Frontend Website chat widget Admin dashboard Lead inbox Booking interface Analytics dashboard Backend Node.js backend PostgreSQL database Prisma ORM Redis for session handling Queue system for background jobs API layer for CRM and calendar integrations AI Layer LLM for conversation generation Intent detection Lead scoring logic Qualification question engine Conversation summarization Follow-up recommendation system Integrations HubSpot Salesforce Zoho CRM Google Calendar Calendly Slack Email WhatsApp Business API How It Works Technically Visitor sends a message through the website chat widget Backend stores the conversation AI detects intent and lead stage AI asks qualification questions based on business rules Lead data is saved in the database AI calculates lead score If qualified, the system books a meeting or sends the lead to CRM Sales team receives a summary with contact details and next steps Business Value This helps businesses: Capture more leads Reply instantly Save sales team time Focus only on qualified prospects Increase booked calls Reduce missed opportunities Improve follow-up quality Convert website visitors into sales opportunities Final Positioning An AI sales agent that talks to your website visitors, qualifies leads, and books meetings with your sales team automatically.
0
18
AgentKit
(1)
Follow
Message
Explore people
