Discover how Kreeda Labs has been making waves capitalizing on the AI technology with its innovative projects.
Case Studies
Solution:
Ayurveda bespoke LLM with Chatbot for end
consumers, trained on 100’s of Client’s
Proprietary Books with Support of Multiple
Languages like Sanskrit, Hindi, English
Features:
• Fast response and 24x7 availability.
• Multi-lingual support.
• Multimodal Input: User can provide input and
receive output in either text or voice.
• Separate Knowledge base of books for every
language to avoid runtime language
translation latency.
• Pipeline for Future Knowledge(Books) training
and enrichment
Fashion Analysis and Trend Forecasting
Founders of Ajio.com - INDIA
Solution:
Platform to foster the Client’s Fashion Research team to Optimize their work and increase efficiency to Analyze the Trends and generate Trend predictions at scale
Features:
•Research and Knowledge enrichment features with web crawlers and social media data gathering bots.
•Image recognition and Analysis leveraging Multimodal Models.
•Content Generation – Text,Image
•Sentiment analysis with Text and Image inputs
•Team Collaboration Boards and Sharing
•Stakeholders – Designer, Forecaster, Consumer
•Trend Analysis Curves with Time Period and Customer Engagements on Social Media Posts
•Innovation and Daily Viral Feeds for research team
Generative AI based Insurance Chatbot to augment the tasks of the human insurance agents as well as optimize the customer engagement and interactions with the intentions to help them buy, decide, know more about the categories of insurance and their plans. Some of the features of the platform are:
-
A conversation experience which is on par with the human agent or even better.
-
Fast response and 24x7 availability.
-
Multi-lingual support.
-
Multimodal Input: User can provide input and receive output in either text or voice.
-
Advanced RAG: Adapt the architecture and hyperparameters of the RAG to adapt to the data better.
-
Use of SOTA elements such as GPT-4 turbo, openai-embedding-3-large, Pinecones serverless vector db. These features have been released not more than 2 months before.
-
CRUD page: A Streamlit and Python based platform to add/update the information in vector database without the manual conversion of the text to embeddings and to avoid reuploading all the information together again.
-
Insurance Case Status: Fetch the application status by providing a valid case id.
-
Lead Generation/ Information Capture: Capture important aspects of the conversation to further process it as an underwriting, leads or downstream use for data analytics.
We proudly introduce the Statistical Assistant Chatbot, a pioneering solution blending artificial intelligence and conversational technology, tailored for the citizens of South Africa.
​
Comprehensive Functionality: Our chatbot is your go-to resource for inquiries and clarifications regarding South African demographics.
With its ability to process both text and voice prompts, it offers detailed statistics and reports in a user-friendly manner.
​
Advanced Interaction: Seamlessly bridging text and speech inputs, our chatbot delivers contextually relevant responses, ensuring a smooth and intuitive user experience.
​
Optimized Infrastructure: Leveraging a Large Language Model and key elements like a vector database and the Next.js serverless platform, we've orchestrated an infrastructure geared towards delivering optimal chat experiences.
​
Continuous Enhancement: Our project entails iterative evolution of both data and models, consistently pushing benchmarks to new heights, guaranteeing the highest standards of accuracy and reliability.
In collaboration with a leading Med-Tech Client, we engineered a state-of-the-art spinal cord segmentation system, revolutionizing the creation of 3D models from MRI scans and X-rays.
​
Enhanced Accuracy through Innovative Processes: Despite limited accessible datasets, our team devised a meticulous process for dataset creation and mask annotation. This innovation significantly boosts accuracy, overcoming industry challenges.
​
Leveraging Cutting-Edge Technology: Harnessing Nvidia's Monai, an advanced medical segmentation model, exemplifies our commitment to employing cutting-edge technology in medical imaging tasks.
​
Unmatched Precision: Our model achieves an exceptional accuracy rate, surpassing 90% for segmentation tasks, validated by a remarkable 0.9 Dice coefficient.
LoadEazy is a Trucking and Logistics Marketplace Full Stack solution with end-to-end delivery, from UIUX, Software Architecture & Development to launch for Logistics marketplace. The solution brings Customers, Transporters and Drivers on one platform to facilitate logistics services for industry grade steel and iron.
​
LoadEazy streamlines and optimizes logistics operations for Heavy Industries by offering a centralized hub where businesses can find and engage with logistics service providers and resources.
Launched in Bengaluru, LoadEazy plans to spread its wings across India soon.
Kreeda Labs has built an AI-powered platform, RobotLawyer.ai which solves the problem of Law approachability for common people.
​
RobotLawyer.ai is an NLP-based OCR Summarization Tool powered by hybrid models and the chatgpt4 framework by Open AI.
​
RobotLawyer.ai will be having access to a various set of documents containing legal information in the form of text files, PDFs, spreadsheets, and so on. The platform treats this as training data and uses a hybrid model over the chatgpt foundational model to learn the context of the documents.
​
The user can login to the platform to seek help on simplifying complex legal documents with the important sections highlighted.
Meet Yipli - A Gamified Fitness Product, making fitness easy and interesting for families. Yipli is a Smart Exercise Mat clubbed with free interactive fitness games and free digital workouts to keep an entire family fit.
​
Kreeda Labs had developed the complete product including hardware Prototype, the App, the Games, the Digital coach as well as the community engagement tools for the Yipli Fitness Platform.
​
Yipli product had been leveraged on extensive technologies like:
Unity, Blender 3D, Photoshop, Figma, Adobe XD
Java, C/C++, Flutter, Node.js, React.js, Firebase Realtime
In the "Rees Rally Racers Multiplayer" project, KreedaLabs embarked on a transformative journey that encompassed diverse aspects of game development.
Our creative efforts materialized as we meticulously developed the game, breathing life into its vibrant world.
A comprehensive overhaul was undertaken, redefining game mechanics, dynamics, and the very environment in which the game unfolds.
This redesign aimed to elevate player engagement and immersion to new heights.
In addition, we conceived and meticulously crafted a fresh game mode "Local Multiplayer". This innovative feature not only injected novelty but also fostered spirited competition among players.
Kecaps is a video entertainment platform that offers content and commercial experiences in a unique way. Kecaps relies on an AI-based recommendation engine for advertisers to bid for ads at the right moment in the videos. The ads are subtle and linked to the e-comm directly.
​
KreedaLabs has developed this Artificial Intelligence AI-based content discovery system aiming to analyze video scenes by identifying content, objects, locations, and people and understand the characteristics in such a way as to deduce value categorizations such as:
•The mood (e.g., cheerful, gloomy, sentimental, energetic, etc.)
•The context (e.g., sports, identifying its type, dialogue, fantasy, action, etc.)
•The places (sea, mountain, city, house interiors, hotels, etc.)
•The colors (prevailing coloring)
•The sound (dialogues, type of soundtrack, type of sounds)
The "HiLo" project is a groundbreaking effort in maritime risk management and safety enhancement. Through innovative data analytics, it aims to prevent high-impact incidents at sea. Divided into three key modules—Journey Overview, Ship Risk, and Port Risk—it equips captains with real-time insights.
By empowering captains with informed decision-making tools, "HiLo" enhances maritime safety, benefiting crew, vessels, and the environment.
In this project, Kreeda Labs team has developed a comprehensive UI/UX solution for the "HiLo" maritime risk management system tailored to ship captains. This encompassed crafting a dynamic dashboard for real-time insights into weather, piracy threats, visibility, and traffic density for optimized route planning.
​
