Freelancers using LangChain in IslamabadFreelancers using LangChain in Islamabad
ProjectsPeople
Armughan Shahid's avatar
Armughan Shahid
proIslamabad, Pakistan
AI SaaS Dev | LLMs, Agents, Voice & Automation | Web, Mobile
New to Contra
AI SaaS Dev | LLMs, Agents, Voice & Automation | Web, Mobile
Cover image for InForm - AI-Driven Physiotherapy App
InForm - AI-Driven Physiotherapy App for Diagnosis, Rehab & Recovery Tracking The core problem it solves: Patients struggle to get timely, structured physiotherapy guidance, while physiotherapists are overwhelmed managing cases, tracking progress, and creating personalized rehab plans manually. Existing systems either lack intelligence or lack control. This product creates a complete digital workflow where diagnosis, communication, and recovery are all connected into a single system. What was built: A full AI-powered physiotherapy platform with mobile apps for patients and a web-based admin system for physiotherapists. Patients begin by submitting structured symptom data through guided questionnaires. Instead of jumping directly to conclusions, the system uses an AI diagnosis engine to analyze patterns and generate internal clinical suggestions. These suggestions are never shown to patients, they act as decision support for physiotherapists, ensuring every diagnosis remains human-approved. Once a case is created, physiotherapists review patient data, validate or override AI recommendations, and communicate directly through an in-app messaging system. Every interaction is tied to a structured case, ensuring context is never lost. The system then moves into recovery management. Physiotherapists create personalized rehab plans with multi-phase programs, exercise libraries, and video guidance. Patients follow these plans inside the app, logging progress, completing KPI-based milestones, and moving through recovery phases in a structured way. A key part of the system is the progress tracking engine. Patients log metrics, complete phase-based goals, and unlock new stages only when criteria are met. Physiotherapists get real-time visibility into adherence, performance, and patient feedback, making the system both trackable and measurable. Alongside this, an intelligent chatbot handles general queries and reduces load on physiotherapists. When confidence is low or cases become complex, the system escalates conversations into structured cases with full context preserved. Technical architecture: Built as a scalable mobile-first system using React Native for patient apps and React.js for the admin panel, with a Node.js/Nest.js backend and Python powering AI-driven diagnosis logic. Data is managed through PostgreSQL, with secure authentication and encrypted communication layers. The AI layer is designed as a support system, not a replacement. It uses LLM-based reasoning to assist diagnosis and continuously improves through feedback loops based on physiotherapist decisions, creating a human-in-the-loop learning system. Deployed on cloud infrastructure (AWS/GCP), with Stripe integration for subscription-based access and a modular architecture designed for future AI expansion. Business model built in from day one: Subscription-based access for patients, with clear pathways to expand into AI-assisted rehab recommendations, outcome analytics, and protocol optimization. The system is designed to evolve into a data-driven recovery platform where every patient interaction improves future treatment accuracy, turning clinical workflows into scalable intelligence.
0
139
Cover image for AI Operations Agent: RAG-Powered Retail
AI Operations Agent: RAG-Powered Retail Intelligence & Task Automation This project was built for large-scale restaurant groups and multi-unit retail operators who manage high volumes of data across dozens or hundreds of locations. Specifically designed for Regional Managers and Operations Directors, the system serves as an enterprise-grade "Digital Consultant" that bridges the gap between fragmented POS/inventory data and daily on-the-ground execution. By transforming millions of rows of restaurant performance metrics into high-priority tasks, it provides a centralized platform for leadership to monitor KPIs, approve AI-suggested corrective actions, and ensure operational consistency across their entire portfolio. 1. What We Built We developed a production-ready Autonomous AI Operations Agent designed to bridge the gap between complex retail data analysis and daily execution. The system acts as a digital consultant for regional managers, transforming raw KPIs into actionable tasks. Analytical AI Chat: A free-form conversational interface where users can query performance data (e.g., "Show me the top 5 worst profitable stores in Istanbul for the last 3 months"). Task Management Dashboard: A structured workflow where AI-suggested actions are automatically logged for manager approval or rejection. Automated Action Logic: The agent uses an "Action Suggestion Map" to identify specific defects (like low audit scores or high food waste) and suggest precise corrective measures. Persistent Memory: Includes both short-term memory for the current chat session and long-term RAG memory to maintain context over time. 2. How We Built It (The Stack) The system was engineered for scalability and reliability using a modern, containerized stack:AI Orchestration: LangGraph was used to manage complex, multi-turn reasoning and agentic workflows. Frontend: React/Next.js 14 for a responsive, real-time user interface. Backend & Data: Node.js paired with a PostgreSQL database capable of handling 1M+ records. LLM Access: Integrated via OpenRouter to allow for flexible model selection and switching. Infrastructure: Fully Dockerized to ensure consistent deployment across environments. 3. Challenges We Faced As the system scaled from prototype to processing millions of records, we encountered several critical engineering hurdles: Response Latency: The initial monolithic prompt architecture led to response times exceeding 60 seconds, far slower than the required "ChatGPT-like" speed. Prompt Verbosity & Errors: Complex questions involving multiple variables caused the LLM to lose focus, leading to "reasoning errors" and incorrect SQL generation. Hallucination Risks: In multi-branch queries, the model occasionally fabricated data points, particularly around manager hours and performance metrics. Context Switching Bugs: The agent sometimes struggled to "let go" of a previous topic, continuing to reference an old store when the user had asked about a new city. 4. How We Solved It We re-engineered the core pipeline to transition from a single, heavy agent into a Modular Multi-Step Architecture: 75% Latency Reduction: By decomposing the main logic into smaller, task-specific nodes, we dropped processing time from 60s down to 15s. Task Decomposition & Specialized Models: We stopped using a "one-size-fits-all" model. Instead, we implemented a router that uses lighter, specialized models for SQL generation and action identification, and flagship models only for final reasoning. Granular SQL Generation: Breaking the metadata analysis into narrow sub-steps eliminated SQL hallucinations. The model now only "sees" the specific schema needed for the current sub-task, ensuring 100% accuracy. 10-Point Testing Protocol: We implemented a rigorous QA protocol that specifically verified bug fixes for context switching, task duplication, and chart coverage before final delivery.
0
149
Cover image for BudgetNest — AI-Powered Personal Finance
BudgetNest — AI-Powered Personal Finance SaaS Most people don't track their finances because the friction is too high. BudgetNest removes that friction entirely, every transaction captured automatically, categorised intelligently, and surfaced through analytics that actually help people make better decisions. The core problem it solves: Manual expense logging fails because people forget, get lazy, or simply don't have time. BudgetNest built an automated capture layer that works across every channel a user already operates in i.e. SMS alerts, bank emails, receipt photos, WhatsApp messages, and voice notes in English and Urdu. The system deduplicates intelligently across all input sources so nothing gets logged twice regardless of how it came in. What was built: A complete AI finance platform with five distinct automated capture modes SMS and email parsing for bank transaction alerts, PDF and image bank statement upload with AI extraction, OCR receipt scanning via camera, a WhatsApp bot that accepts text, images, and voice notes, and multilingual voice input for manual cash payments. Every transaction flows through an LLM-powered categorisation engine that auto-assigns categories and subcategories, recognises vendors, and learns from behaviour over time. Beyond capture, the system includes smart budgeting with AI-driven suggestions based on spending patterns, subscription detection for recurring transactions, shared expense and split-bill tracking, fraud detection for unusual transactions, and forecasting that projects deficit against income. Dashboards surface everything through charts, trend lines, and weekly and monthly summaries. Technical architecture: React Native across iOS and Android, Node.js and FastAPI backend, PostgreSQL and MongoDB, AWS infrastructure with EC2, S3, and RDS, Python-based NLP and OCR pipeline using Transformers and Tesseract, Twilio WhatsApp integration, Gmail API for email parsing, and Firebase for push notifications. Business model built in from day one: Freemium with premium automation features, B2B white-label capability for microfinance institutions and NGOs, and the OCR and SMS parsing logic architected as standalone APIs for third-party licensing meaning the AI layer has revenue potential independent of the consumer app.
1
175
Cover image for Share the Light - Role-Based Tutoring Platform
Share the Light - Role-Based Tutoring Platform
0
4
LangChain(3)
Khakan Hayder's avatar
Khakan Hayder
Islamabad, Pakistan
Framer Expert
$1k+
Earned
1x
Hired
5.0
Rating
8
Followers
FRAMER_EXPERT

expert

Framer Expert
Cover image for A quick reel of recent
A quick reel of recent client work delivered through VConekt. Each project went from blank canvas to live, conversion-ready Framer site: covering SaaS landing pages, fintech dashboards, AI product launches, local service businesses, and personal portfolios. What I bring to every project: • Official Framer Expert (verified) • Strategy-first design, not just pretty pixels • CMS, animations, responsive across every breakpoint • Clean handoff or full ongoing management Available for new projects starting June. Let's build.
1
66
Cover image for This is something, I really
This is something, I really liked to work on... https://www.framer.com/community/marketplace/templates/peakcoaching/
0
5
Cover image for Amble : Premium Framer Template
Amble : Premium Framer Template (free for a limited time) I design and build polished marketing websites in Framer, and Amble is my latest template: a refined, editorial layout crafted for agencies, studios, and founders. It ships with 8 ready pages, a CMS powered blog, smooth scrolling, full responsiveness, and unique SEO on every page, all built on a clean, reusable component system that's easy to make your own. For one week, I'm offering it free so more teams can launch something they're proud of. Grab it here: https://amblee.framer.website/ Need a custom site or want Amble tailored to your brand? I'm available for projects. Let's talk 👉 khakan@vconekt.com (mailto:khakan@vconekt.com)
0
16
Cover image for Diamant Versatile
Diamant Versatile
0
5
LangChain(1)
Shehzada Ammad Ali's avatar
Shehzada Ammad Ali
Islamabad, Pakistan
Full Stack Developer| n8n | Framer| Kajabee| Webflow Expert
1x
Hired
46
Followers
Full Stack Developer| n8n | Framer| Kajabee| Webflow Expert
Cover image for AI-Powered News Research Tool | Streamlit Web App This proje...
AI-Powered News Research Tool | Streamlit Web App This project is a fully functional AI-driven news analysis tool built using Streamlit, designed to help users extract valuable insights from online news articles. Key Features: Accepts a news article URL as input Uses Google Generative AI to analyze and understand the content Splits and processes large articles using Recursive Text Splitting Embeds content using Google Embedding Models Indexes data with FAISS for efficient semantic search Allows users to ask natural language questions, returning accurate answers along with cited sources
1
96
Cover image for Q&A Chatbot with CSV-Powered Knowledge Base This project is ...
Q&A Chatbot with CSV-Powered Knowledge Base This project is an intelligent Question and Answer Chatbot built using LangChain, Google Gemini AI, FAISS, and Streamlit. It allows users to interact with a chatbot that pulls highly accurate answers directly from a CSV-based knowledge base.
1
101
Cover image for 🚬 New Project: Migratory Story Excited to kick off a new Sh...
🚬 New Project: Migratory Story Excited to kick off a new Shopify project for Migratory Story, a premium cigar store. Services: Shopify setup, UI/UX design, Figma prototypes, ecommerce management, and sales-driven content. #Shopify #Ecommerce #PremiumCigars #UIUXDesign #Figma #DirectResponseMarketing
1
157
Cover image for Human Resource Intelligence Chatbot with SQL-Driven Data Vis...
Human Resource Intelligence Chatbot with SQL-Driven Data Visualization I developed an AI-powered HR chatbot that transforms natural language queries into actionable insights through SQL-driven analytics and automated visualizations. How It Works: HR asks questions in plain English AI generates/executes SQL queries Processes workforce data Outputs structured results or charts Key Features: Natural language to SQL conversion Real-time workforce analytics Automated data visualization Scalable integration with HR databases Extendable to other chart types
2
191
LangChain(2)
Farhan Khan's avatar
Farhan Khan
Rawalpindi, Pakistan
AI Specilist, AI Automation, Chatbots, Business, Workflow
New to Contra
AI Specilist, AI Automation, Chatbots, Business, Workflow
Cover image for Automated Multi-Agent AI Support &
Automated Multi-Agent AI Support & Lead Triage Pipeline Are your high-ticket clients waiting hours for an email response? This intelligent multi-agent n8n workflow instantly screens, analyzes, and responds to customer emails in real-time, utilizing advanced RAG (Retrieval-Augmented Generation) to deliver human-like support instantly. Project Overview: This is an enterprise-grade AI automation system designed to eliminate manual customer support queues. Instead of simple auto-replies, it uses a multi-agent routing structure combined with a dynamic knowledge base to handle complex inquiries autonomously. How It Works (Under the Hood): Instant Inbound Triage: A Gmail Trigger catches incoming emails instantly, extracting raw content for processing. AI Intent Classification: An initial OpenAI model acts as a gatekeeper, analyzing the email to determine if it is a valid customer support request or irrelevant noise. Conditional Routing: An advanced router splits the path: non-support emails receive a polite automated Telegram update, while actual support tickets are routed to the main AI engine. Context-Aware AI Agent: The core Customer Support Agent is equipped with an OpenAI Chat Model, conversational memory, and a custom Vector Store Tool. Pinecone RAG Integration: The agent queries a Pinecone Vector Database (powered by OpenAI Text Embeddings) to fetch real-time, accurate company documentation and context, eliminating hallucinations. Automated Action & Response: Once the resolution is drafted, the system automatically creates a draft in Gmail for review and sends an instant internal notification via Telegram. Why This Wins Clients (The Value Pitch): Zero Hallucinations: Connected to a live vector database (Pinecone) so the AI only speaks from approved company data. Reduced Overhead: Cuts down customer support response times from hours to under 60 seconds. Production-Ready Architecture: Designed with modern n8n AI sub-nodes, structured tools, and modular scaling capabilities.
1
3
150
Cover image for E-Assistant is an autonomous AI
E-Assistant is an autonomous AI shopping agent designed to streamline the consumer decision-making process. By simultaneously querying multiple e-commerce platforms, it utilizes a proprietary value-ranking algorithm to provide real-time product comparisons based on price, rating, and review volume.
0
31
Cover image for Automated Video Dehazing & Atmospheric
Automated Video Dehazing & Atmospheric Haze Simulation System 🚀 Project Overview This advanced Computer Vision project is designed to address visibility challenges in adverse weather conditions. The system features a dual-module architecture: it can synthetically inject realistic atmospheric fog/haze into crystal-clear video streams for dataset generation, and conversely, restore heavily degraded, foggy videos into crisp, high-visibility outputs in real-time. 🛠️ Core Functionality & Modules Module 1: Atmospheric Haze Simulation Purpose: Generates synthetic datasets to train and benchmark object detection models (like YOLO) for bad weather conditions. How it works: Implements mathematical scattering models to calculate depth maps and overlay a realistic layer of dense fog or smoke over clean video frames. Module 2: Real-Time Video Dehazing Purpose: Restores clarity and vivid color to video streams captured in low-visibility environments. How it works: Leverages physics-based Computer Vision algorithms (such as Dark Channel Prior - DCP) or Deep Learning frameworks to estimate atmospheric light, eliminate transmission noise, and reconstruct the scene's original contrast. 🎯 Use Cases & Applications Autonomous Vehicles: Enhances the sight and reliability of self-driving car sensors in dense fog. Smart Surveillance (CCTV): Improves security monitoring and facial recognition accuracy under harsh outdoor weather. Drone Navigation: Aids aerial drones in safely navigating through smoke, dust storms, or low-lying clouds. 💻 Tech Stack Used Language: Python Libraries: OpenCV, NumPy, Matplotlib, PyTorch / TensorFlow (if deep learning was applied) Concepts: Image Processing, Atmospheric Scattering Models, Feature Restoration, Video Pipeline Optimization
0
42
Cover image for Real Time SMS Spam Detection/Classification
Real Time SMS Spam Detection/Classification System
1
81
LangChain(1)
syed zain hasan's avatar
syed zain hasan
Rawalpindi, Pakistan
End-to-End Blockchain & AI Solutions Expert
7
Followers
End-to-End Blockchain & AI Solutions Expert
Cover image for StonAI
StonAI
1
8
Cover image for TMInspector- Trademark Infringement Solution
TMInspector- Trademark Infringement Solution
1
6
Cover image for Pickle Arcade- Cardano Multiplayer Gaming Platform
Pickle Arcade- Cardano Multiplayer Gaming Platform
1
3
Cover image for JobParty: GenAI based CV Builder and Career Coach
JobParty: GenAI based CV Builder and Career Coach
1
3
LangChain(2)
Arham Malik's avatar
Arham Malik
Rawalpindi, Pakistan
Backend & AI engineer who ships systems fast and scale.
New to Contra
Backend & AI engineer who ships systems fast and scale.
Cover image for Avenix is basically the biggest
Avenix is basically the biggest thing I built solo. It’s an enterprise AI platform that actually feels smart and fast. I used LangChain agents, real time WebSockets, and a whole lot of TypeScript and Python to make it work. The whole idea was to let businesses have real conversations with their data, with memory and context that doesn’t break. I delivered it three months early, and the client was over the moon. Queries got 40% faster. Honestly, it was a beast, but I loved every minute. For more details, please visit: https://arham-nexus.vercel.app/work/avenix
0
55
Cover image for This is a classic enterprise
This is a classic enterprise academic portal, but built solid. It uses  Web Forms, ASP.NET, C#, and SQL Server. The main challenge was role based access control because you have admins, faculty, TAs, and lab demonstrators all needing different permissions. I designed a three tier architecture with stored procedures and optimistic concurrency control so data doesn’t get messed up when people edit at the same time. Over 200 users per semester use it for task assignments and progress tracking. It’s not flashy, but it works perfectly. For more details, please visit: https://arham-nexus.vercel.app/work/talabportal
0
30
Cover image for Skill Swap is a peer
Skill Swap is a peer to peer learning platform where you can be both a student and a teacher. I built the backend with Java Spring Boot, real time chat with WebSockets, and video calls with WebRTC. The matching algorithm finds people based on skills, ratings, and availability. You can switch roles anytime. It also has a trust based review system so fake reviews don’t ruin it. This was an MVP, but it proved that real time learning communities can work really well. For more details, please visit: https://arham-nexus.vercel.app/work/skillswap
0
48
Cover image for This is a native Android
This is a native Android app I built with Kotlin and Firebase. It connects tourists with verified local guides in real time. You open the app, see nearby guides, book a tour, and then you can track each other’s location live using Google Maps. I also added offline support because travel doesn’t always have perfect internet. Over 100 active users ended up using it,during the MVP. The best part? Battery usage was optimized so your phone doesn’t die halfway through the day. Really proud of how smooth it turned out. For more details, please visit: https://arham-nexus.vercel.app/work/raaheraast
0
51
LangChain(1)