Google PaLM Applications

Browse applications built on Google PaLM technology. Explore PoC and MVP applications created by our community and discover innovative use cases for Google PaLM technology.

Retriever AI

Retriever AI is an innovative software solution that leverages cutting-edge artificial intelligence technology to revolutionize the way users interact with their Windows operating systems. By leveraging the capabilities of OpenAl's Whisper Automatic Speech Recognition (ASR) system and ElevenLabs' advanced interaction the application delivers a transformative user experience. Users can interact with their computers using natural spoken language, receive auditory feedback, and carry out tasks without the traditional visual interfaces. At its core, Retriever AI is powered by advanced machine learning algorithms that enable it to understand and respond to user commands effectively. With a simple "Start" command, users can invoke Retriever AI to assist them in navigating their system, opening applications, searching for files, and much more. It is like having a personal assistant dedicated to making your computer interactions more efficient and enjoyable. The software is designed with a user-friendly interface that is easy to start and stop, and it's designed to be almost hands-free from the keyboard. Its design is meant for the visually impaired and blind, and it's geared toward being able to complete normal functions using natural language. In a digital world where efficiency and user experience are of utmost importance, Retriever AI serves as a valuable tool for enhancing productivity, simplifying tasks, and creating a more intuitive interaction between users and their Windows systems even if you aren't visually impaired or blind. Whether you're a professional looking for a smarter way to navigate your workspace, a student aiming for better efficiency, or just a casual user hoping to get more out of your system, Retriever AI is designed to meet your needs.


VBCST voice based customer support tool

VBCST is a voice-based customer support tool that can talk to customers It can be used to manage business queries and replace boring customer agents at your business. VBCST is powered by a large language model such as palm 2 that has been trained on a massive dataset. This allows VBCST to understand customer queries and provide accurate and helpful responses. VBCST can also access metadata about the customer, such as their name, contact information, and purchase history. This information can be used to personalize the customer experience and provide more relevant support. VBCST is a cost-effective way to improve customer support. It can be used to handle a large volume of calls, freeing up human agents to focus on more complex queries. VBCST can also be used to provide 24/7 support, which can be a valuable asset for businesses that operate in multiple time zones. VBCST is easy to use. It can be integrated with existing customer support systems, and it does not require any special training. VBCST can be used by businesses of all sizes, and it is a cost-effective way to improve customer satisfaction. Here are some of the benefits of using VBCST: Increased customer satisfaction: VBCST can provide accurate and helpful responses to customer queries, which can lead to increased customer satisfaction. Reduced costs: VBCST can help businesses to reduce the cost of customer support by handling a large volume of calls. Improved efficiency: VBCST can help businesses to improve the efficiency of their customer support by providing 24/7 support and by freeing up human agents to focus on more complex queries. Personalized customer experience: VBCST can access metadata about the customer, such as their name, contact information, and purchase history. This information can be used to personalize the customer experience and provide more relevant support. Here for the project purpose we have made a customer support tool for tesla company and it can be used in different companies too.


Research assistant

This project revolves around the development of a research assistant using the Google Vertex AI Palm2 platform. The aim is to streamline the process of searching for and accessing academic papers from Google Scholar, providing researchers with a user-friendly and efficient tool. The research assistant is implemented as a Streamlit application, allowing users to input their search specifications and navigate through Google Scholar seamlessly. One of the key features of the research assistant is its automatic scraping functionality. Once the user provides their search criteria, the application scours Google Scholar across multiple pages, retrieving relevant papers. The scraped papers are then organized into a comprehensive dataframe, providing researchers with a structured overview of the available literature. Additionally, the application also selects and provides downloadable PDF versions of the papers, making it convenient for users to access and read the full content. To further enhance the capabilities of the research assistant, it integrates with Google Vertex AI and Langchain. Google Vertex AI is a powerful machine learning platform that enables users to leverage advanced AI models and tools. By integrating with Vertex AI, the research assistant allows researchers to create a knowledge base from the downloaded papers, enabling them to extract insights and answer questions related to the content. Langchain, another crucial component, provides additional functionality for knowledge extraction. It offers a range of AI models and tools specifically designed for language processing and analysis. Integrating Langchain with the research assistant expands its capabilities, allowing researchers to delve deeper into the papers and extract valuable information.



GhaGeni is developed using Web Assembly (WSAM) and Flutter. It incorporatesGoogle Cloud's large language model (Palm2 )for intelligent conversations. The user interface is designed for a seamless experience. GhaGeni utilizes a Redis database for efficient data management and prompt storage. The core idea behind GhaGeni is to provide users with a fresh perspective and help them overcome mental roadblocks by asking thought-provoking questions. It recognizes that individuals often struggle to remember crucial details or fully explore different angles when working on complex tasks. GhaGeni steps in to bridge that gap and facilitate a more fruitful and efficient problem-solving experience. When a user engages with GhaGeni, they begin by stating their objective or desired outcome. It could be anything from designing a website to planning a marketing campaign. GhaGeni then responds with a set of three questions related to the user's objective. These questions are carefully crafted to prompt the user to consider various aspects, explore different angles, and uncover hidden insights. As the user responds to each question, GhaGeni dynamically generates the next set of three questions based on their previous answers. This iterative process guides the user down a path of continuous discovery and exploration. The questions become increasingly tailored and specific to the user's unique needs, enabling them to dive deeper into their project and consider aspects they may have overlooked. Whether you're a professional working on a complex project, a student grappling with an academic task, or an individual looking for inspiration and guidance, GhaGeni is there to support you every step of the way. Embrace the power of interactive conversations and unlock your true genius with GhaGeni.

RedisPaLMLangChainText Generation Web UI

NeuroGuru - AI Virtual Learning Assistant

NeuroGuru is a Vertex - Generative AI-powered educational platform designed to help users around the globe learn about artificial intelligence (AI) in their native language. Leveraging the advanced capabilities of models like PaLM2, Codey, NeuroGuru provides a personalised and interactive learning experience that adapts to each individual's learning pace and style. The platform covers a wide range of AI topics including Machine Learning, Deep Learning, Natural Language Processing, and Reinforcement Learning. It's not just a passive learning platform - NeuroGuru is designed to be interactive, engaging users with quizzes, hands-on coding exercises, and advanced topic explorations based on their interests and goals. One of the standout features of NeuroGuru is its adaptive learning capability. Based on the learner's progress and responses, NeuroGuru adjusts the learning content, making recommendations for what to learn next. It also includes a community feature for learners to connect, discuss, and collaborate on AI topics. NeuroGuru also emphasizes convenience and accessibility, with voice assistant integration and cross-platform support ensuring a seamless learning experience across computers, tablets, and smartphones. Whether you are a beginner just stepping into the world of AI, or an experienced professional wanting to keep up-to-date with the latest AI advancements, NeuroGuru is your comprehensive, go-to platform for AI education.

AI Wonder Girls
PaLMCodeyGenerative AI Studio

Complementary Opposing Views - CON

News platforms are often driven by profit and incentivized to maximize user attention through echo chambers and increasingly polarizing news. This phenomenon has been observed in various media platforms, such as cable news during US elections, trending controversial videos on YouTube and TikTok, and the spread of fake news. To address these challenges, Complementary Opposing News (CON) provides readers with a diverse range of news sources and challenges them to engage with opposing views. By presenting readers with complementary opposing viewpoints, CON aims to promote critical thinking and reduce polarization. Furthermore, CON plans to leverage Google Vertex AI to make playing devil's advocate a pro-social activity for opposing viewpoints to complement each other's understanding of the world instead of tearing each other down. By using these technologies, CON aims to promote critical thinking by encouraging readers to play devil's advocate as they're exposed to opposing viewpoints in steel-manning arguments. CON aims for a more informed and less polarized society, one chrome extension download at a time. Down the road, CON hopes to become more than just an isolated chrome extension and expand internationally. By leveraging Google Vertex AI and Firebase for production-grade deployment to scale the solution to many more users and support multiple languages, reducing international conflict. Furthemore, CON is optimistic in using blockchain technologies to incentivize communities for truth-seeking and become a platform for Socratic discourse. Con's frontend is React, Typescript, Chakra UI, and backend is FastAPI and GCP. Con uses langchain and Google Vertex AI for ML models for sentiment analysis of articles and text generation and conversational assistance. CON hopes to expand the platform to offer data analysis and visualization to help users engage with a broad spectrum of opinions and determine the credibility and believability of different news.

PaLMLangChainText Generation Web UIModel Garden


ntroducing TalkToMe, a groundbreaking web application that revolutionizes the way we engage with podcasts, books, and various forms of documentation. Gone are the days of passive consumption; now, we enter a realm of interactivity and immersion. TalkToMe employs cutting-edge technologies, harnessing the power of advanced Large Language Models, Speech-to-Text, and Vision models provided by Google Cloud Services. This amalgamation of state-of-the-art AI enables us to deliver an unparalleled user experience. Imagine effortlessly uploading audio files, books, PDFs, or any content of your choosing, triggering the creation of a dynamic ChatSession. Our web-app embarks on an intellectual journey through the depths of your uploaded material, extracting its very essence and comprehending its context. This deep understanding empowers TalkToMe to provide you with insightful responses to your queries. It's an interactive symphony. Utilizing intuitive speech interaction, you can actively engage with the ChatSession, asking questions that penetrate the core of the content. Prepare to be amazed as TalkToMe offers concise and informative answers, guiding you on an intellectual odyssey. But TalkToMe doesn't stop there; its capabilities transcend conventional boundaries. Summarization becomes effortless, distilling the essence of lengthy material into digestible nuggets of wisdom. General comparisons unveil hidden truths, shedding light on similarities and disparities. The world becomes your intellectual playground as TalkToMe empowers you to embark on an all-encompassing exploration of knowledge. Unlock the true potential of your chosen materials with TalkToMe, transforming them into interactive companions on your journey of discovery. Immerse yourself in a realm where learning and enjoyment converge, where the boundaries between content and consumer dissolve. Embrace the future of interactive content consumption and join us as we rewrite the rules of engagement.

application badge
Model GardenVercelChirpLangChainGenerative AI StudioPaLM


Our idea is to develop a Personal Growth Motivation App that serves as a comprehensive platform for individuals seeking emotional relief, spiritual enrichment, and personal growth. The app aims to provide users with the necessary resources, guidance, and tools to embark on a transformative journey of self-discovery and self-improvement. The app's core features revolve around two key components: the "Detailed Question Stage" and the "Interactive Diary." In the "Detailed Question Stage," users can choose from a range of themes such as Finding Balance, Practicing Self-Care, Exploring New Things, and Gratitude and Kindness. Within each theme, users are presented with thought-provoking questions specifically tailored to their chosen topic. These questions are designed to facilitate deep introspection, encourage personal growth, and foster self-awareness. The app's underlying large language model analyzes user responses and provides personalized guidance and further questions based on their answers. This interactive process allows users to delve deeper into their thoughts, emotions, and motivations, enabling them to gain a greater understanding of themselves and their desires for personal growth. Our Personal Growth Motivation App aims to empower individuals on their personal growth journeys, helping them find emotional balance, discover new perspectives, cultivate self-care practices, express gratitude and kindness, and ultimately achieve a greater sense of fulfillment and well-being.



Our groundbreaking project aims to bridge the mental health gap in low-access regions and underdiagnosed populations of developing countries. With limited access to traditional healthcare services, many individuals struggle silently with mental health challenges, often going undiagnosed and untreated. Our mission is to provide a lifeline of support and guidance through our innovative psychiatric chatbot. Leveraging the power of artificial intelligence and natural language processing, our chatbot serves as a virtual mental health companion, offering a safe and confidential space for individuals to express their thoughts, feelings, and concerns. It engages in empathetic conversations, actively listening and providing evidence-based guidance, empowering users to understand their mental health better. Tailored to the specific cultural and linguistic nuances of the target regions, our chatbot provides personalized assistance, making mental health resources accessible and understandable to those who need them most. Through a user-friendly interface, individuals can access self-help tools, coping strategies, and educational content to enhance their mental well-being. Our project operates on the principle of inclusivity, ensuring that individuals from all walks of life, regardless of socio-economic status or geographical limitations, have access to crucial mental health support. By leveraging digital technology, we aim to break down barriers, reduce stigma, and promote early detection and intervention for mental health issues. Together, let us empower individuals in low-access regions, providing them with the support and resources they need to lead mentally healthy lives. Join us on our journey to make mental well-being a universal right, reaching the underserved populations and bringing hope and healing to those who need it the most.

Generative AI StudioPaLM


Storyboard is a web-based app that empowers users to effortlessly create compelling narratives. Leveraging cutting-edge technology, PaLM2 for Text and EfficientNetV2, it transforms uploaded text and image files into immersive storylines. Users begin by uploading their text or image files, which serve as the foundation for story creation. Text files are processed using PaLM2 for Text. For image files, EfficientNetV2 analyses the uploaded images to extract key features. These features seamlessly integrate into the story generation process, adding depth to the narratives. Through prompt engineering techniques, user inputs are effectively incorporated, ensuring personalised and coherent narratives. PaLM2's natural language generation capabilities produce captivating and authentic stories. Storyboard allows users to select the genre of the story they want the AI to create. Whether it's a thrilling mystery, heartwarming romance, or epic fantasy, users can tailor their storytelling experience to match their preferences. This genre selection ensures the AI-generated stories align with the user's interests, providing an enjoyable and personalized experience. The motivation behind Storyboard is rooted in the profound impact storytelling has on human connection and personal growth. For centuries, storytelling has allowed us to share experiences, explore perspectives, and cultivate empathy. However, not everyone has the time, skill, or resources to create engaging narratives. Storyboard aims to break down these barriers, enabling a broader audience to experience the transformative power of narratives. By providing an intuitive platform that harnesses the capabilities of advanced AI models, Storyboard empowers users to become storytellers in their own right. Whether it's for personal reflection, creative expression, or entertainment, Storyboard opens up a world of storytelling possibilities, fostering personal growth and connection through the magic of narratives.

Asparagus Taco
PaLMModel Garden


Communication barriers and challenges exist for individuals who are deaf, hearing-impaired, or have difficulty making phone calls. These individuals may face limitations in understanding spoken language, maintaining focus, managing distractions, and effectively participating in phone conversations. Additionally, introverts may experience discomfort or anxiety when engaging in verbal communication. These factors hinder inclusivity, independence, and effective communication for these user groups. Solution: Our product, ConvoAI, offers a transformative solution to address these challenges. By harnessing the power of AI voice recognition, content generation, and real-time assistance, ConvoAI enables individuals to make phone calls with ease, confidence, and enhanced communication capabilities. The key features and benefits of ConvoAI include: Content Generation and Recommendations: ConvoAI generates AI-powered responses, prompts, and suggestions, reducing the need for constant input from the user and promoting engaging and smooth conversation flow. Personalized Experience: ConvoAI can be tailored to individual preferences, including language settings, visual cues, and content generation options, providing a personalized and comfortable communication environment. Time Management and Summaries: ConvoAI helps users manage call duration, offers time-related prompts, and provides post-call summaries of key points, action items, and important details discussed. By leveraging these powerful features, ConvoAI empowers deaf, hearing-impaired, introverts, and other individuals who face communication challenges to engage in phone conversations with confidence, independence, and improved comprehension. Our product enhances inclusivity, fosters effective communication, and ultimately enriches the lives of users by breaking down communication barriers.

Just Do AI
application badge


AIT-DiagramCodeX is a groundbreaking platform designed to redefine software development. By leveraging advanced AI technology, it seamlessly transforms natural language into diagrams, and further, into source code. This unique solution streamlines the workflows of developers, engineers, and data scientists, enhancing productivity and efficiency. The platform hosts a unique four-part user interface, which accepts natural language descriptions and uses AI to convert them into comprehensible PlantUML diagrams. Furthermore, it generates a clear PNG image of the diagram for easy communication and sharing. In a revolutionary step, AIT-DiagramCodeX also provides ready-to-use Go app source code directly from the diagrams. Hosted on Google Cloud Platform's CloudRun, AIT-DiagramCodeX promotes a collaborative environment that is accessible and shareable, creating a harmonious space for developers to work together. It employs smart pipeline systems and advanced technology to ensure a smooth, optimized transition from natural language to diagram, and subsequently, to code. AIT-DiagramCodeX is more than just a toolβ€”it's a transformative solution designed to supercharge productivity in large enterprises. Its integration of Google's Vertex AI PaLM for accurate conversions between natural language, diagrams, and code sets it apart from other platforms in the market. Our vision with AIT-DiagramCodeX is to redefine the way we understand, share, and create in the digital world. We are not only creating a software development tool but shaping a revolutionary journey in the digital landscape. Join us as we move towards a future where technology seamlessly converges with human creativity.

PaLMText Generation Web UI

Vectex AI GenAI with memory

Vectex AI is a project that seamlessly fuses VectorStore, an advanced vector search library, with Google's Vertex AI, setting a new precedent for AI-enhanced, fact-based conversations. The project's nucleus resides in the innovative use of VectorStore, a tool well-primed for handling high dimensional data. The VectorStore database, stored within a Google Cloud bucket, serves as an extensive reservoir of information. It powers the project's memory retrieval capabilities, allowing for comprehensive and factual responses to user queries. The integration of Vertex AI complements the project's ambition. Google's versatile machine learning platform ingests the user's conversational prompts, and leveraging its pre-trained model, executes a search process within VectorStore. The 'cosine' distance metric and a cap of '20' results optimize this search, ensuring the system retrieves the most relevant data for each query. The marriage of Vertex AI's knowledge retrieval with VectorStore's memory capacities creates a powerful synergy. It allows the AI to engage in conversation, while simultaneously accessing and integrating factual knowledge. The result is a dialogue that's not only intelligent but contextually enriched and accurate. The project is encapsulated within a sleek web UI, courtesy of Vue.js and Tailwind CSS. This vibrant, user-friendly interface houses the Vertex AI-VectorStore fusion, offering an engaging platform for users to experience these enhanced AI dialogues firsthand. The UI's dynamically updated background image, fetched directly from the asset directory, adds a captivating visual touch, making the dialogue process more immersive.

application badge

Google Vertex is the SBOM

Common Vulnerabilities and Exposures (CVEs) are a standardized way of cataloging vulnerabilities in software. When a security vulnerability is discovered, it can be assigned a CVE identifier to ensure that information about the vulnerability can be shared and referenced across different platforms and organizations. CVEs are currently maintained as freeform text, but will soon be maintained in JSON format. A Software Bill of Materials (SBOM) is a document that provides a comprehensive inventory of software components and dependencies. It is a structured list that includes information about the software components, libraries, frameworks, and other third-party software used in the development or deployment of a software product. SBOMs can be automatically generated from files that define build information for software, e.g. POM files for Maven, requirements.txt for Python, etc. It would be extremely useful to be able to automatically scan SBOMs to see whether the associated software packages are vulnerable to the CVEs. Google Vertex's Palm turned out to be capable of automatically finding a list of affected versions based on the CVE descriptions. For example, there was a major vulnerability in Apache Log4J in late 2021. Let's take a look at the version information in the CVE ( "Apache Log4j2 versions 2.0-beta7 through 2.17.0 (excluding security fix releases 2.3.2 and 2.12.4) are vulnerable to a remote code execution (RCE) attack...." That makes sense to a human, but it's not conducive to automatic processing. Google Vertex was able to generate a specific list of Log4J versions that met the criteria in the description. It also shows potential for automatically extracting additional information from text descriptions in CVEs. With this information, we can automatically read an SBOM and determine whether the software is vulnerable to a particular CVE.

SBOM is the Bomb

TACO - Terms and Conditions Oracle

Introducing TACO, a powerful app powered by Google Palm designed to assist users in answering questions about the terms and conditions of various renowned companies. Whether you're curious about the policies of Google, Meta (formerly Facebook), Apple, Uber, Lyft, Reddit, Snapchat, or Twitter, TACO has got you covered. With TACO, users can conveniently access a comprehensive database of terms and conditions from these prominent companies. By leveraging its advanced language processing capabilities, the app enables users to ask specific questions or seek clarification on any confusing clauses within the terms and conditions of these platforms. TACO's user-friendly interface allows for seamless navigation, ensuring a hassle-free experience for users seeking accurate and reliable information. Whether you need to understand data usage policies, privacy settings, content guidelines, or any other aspect of these companies' terms and conditions, TACO delivers concise and precise answers. By harnessing the power of the LLM technology, TACO provides users with an efficient way to navigate the complex world of legal agreements. With this app at your fingertips, you can make informed decisions, protect your rights, and stay up-to-date with the policies of the companies shaping the digital landscape. TACO empowers users with knowledge, allowing them to understand and navigate the intricacies of the terms and conditions set forth by Google, Meta, Apple, Uber, Lyft, Reddit, Snapchat, and Twitter. Say goodbye to confusion and uncertaintyβ€”TACO is here to simplify the world of legal jargon and help you stay informed in the digital age.

Prompt Paladins
application badge


Captionize is a cutting-edge AI solution that automates the generation of video descriptions, empowering content creators on YouTube to enhance their productivity, expand their reach, and unlock new revenue opportunities. By harnessing the power of artificial intelligence, Captionize streamlines the creation of video descriptions, saving creators valuable time and providing them with a competitive edge in the digital landscape. YouTube content creators often struggle with crafting engaging video descriptions, limiting their ability to focus on quality content and channel growth. Manual creation is time-consuming and can result in inconsistent or subpar descriptions that hinder outreach efforts and reduce audience discovery. Leveraging advanced AI algorithms, Captionize automatically generates compelling video descriptions. By analyzing the transcript of the video, Captionize creates informative and engaging descriptions tailored to maximize SEO performance, ensuring higher search rankings, increased organic traffic, and improved visibility on YouTube. Captionize presents a compelling business opportunity for both the product and its users. By saving time and offering unique benefits, Captionize is poised to capture a significant market share, providing substantial profits and success to content creators in the growing industry. In conclusion, Captionize revolutionizes video descriptions for YouTube content creators, offering a time-saving, AI-driven solution that optimizes SEO, expands reach, and unlocks new revenue opportunities. With its unique features and benefits, Captionize is well-positioned to thrive in the content creation market, delivering significant profits and success for both the product and its users.

The Vertex Titans
application badge
Generative AgentsPaLMModel GardenText Generation Web UI

StoryGen - Empowering Moral Education through AI

StoryGen represents a groundbreaking initiative poised to revolutionize moral education and character development for children globally. Our mission is to promote global moral education by leveraging artificial intelligence to adapt ancient fables from diverse cultures. In our interconnected world, it is vital to instill strong moral values while embracing the diversity of global cultures. Traditional fables have long been revered for their wisdom. However, by expanding our repertoire to include fables from various ancient traditions, we have an opportunity to create a truly inclusive and impactful educational experience. Our goal is to adapt these fables using AI techniques, ensuring they resonate with children worldwide. Key Features: Cultural Adaptation: StoryGen employs AI technologies to adapt fables, transcending cultural boundaries. For example, Panchatantra fables can be reimagined with western characters, enabling children in Western countries to enjoy and appreciate Indian wisdom. Similarly, fables from Western cultures can be adapted to resonate with children in other regions. This approach promotes cultural exchange and understanding. Age-Appropriate Content: StoryGen dynamically tailors the complexity and vocabulary of the stories to suit the developmental stage of the target audience. Younger children receive fables with simpler language and themes, while older children engage with more nuanced and thought-provoking narratives. Ethical Lessons and Moral Values: StoryGen carefully selects fables that promote positive values, critical thinking, empathy, and character development. E.g. honesty through "The Boy Who Cried Wolf" and gratitude in "The Lion and the Mouse." These lessons are universally applicable and resonate with children from different cultural backgrounds. Language and Communication Skills: StoryGen enhances language and communication skills through engaging stories. Example Content:

StoryGen - Empowering Children
PaLMChirpModel Garden

Pdf 2 Bot

Introducing Pdf2Bot, an innovative web application designed to revolutionize the way we interact with documents. With Pdf2Bot, users can effortlessly create a dynamic chatbot that seamlessly extracts information from uploaded PDF files and provides accurate answers to a wide range of queries. What sets Pdf2Bot apart is its unique ability to generate response At the core of Pdf2Bot lies the powerful text bison model from VertexAI, a cutting-edge language processing technology. This sophisticated model is specifically trained to understand and analyze textual content, enabling Pdf2Bot to comprehend the uploaded PDF documents with remarkable accuracy. Whether it's a research paper, a technical manual, or a legal document, Pdf2Bot's text bison model can handle diverse types of content, making it a versatile tool for various industries and purposes. Using Pdf2Bot is a breeze. Users simply upload their desired PDF file through the user-friendly web interface. The intelligent backend of Pdf2Bot swiftly processes the document, extracting and organizing the relevant information into a structured format. Once this initial step is complete, the magic begins. Pdf2Bot's chatbot functionality comes into play, as it leverages the processed content to provide intelligent responses to user queries. The chatbot is designed to understand natural language and can handle questions of varying complexity. Whether it's a specific fact, a concept explanation, or a request for further details, Pdf2Bot's chatbot can handle it all of them

application badge
PaLMChromaLangChainChirpText Generation Web UI


Today, we are presenting an idea which can help in routing the Chat/Voice interaction in the contact center to the right skilled agent. As we all know Contact center plays a bigger role in any organization when providing the best customer service to the users. Ideally, in any digital interaction in a contact center ecosystem, like voice, chat or email involves a virtual agent which is the first layer to interact with the customer. VA tries to welcome the user and understand the Intent. Based on the intent, query the VA tried to have the conversation. If the Intent is complex we might need human assistant to look upon, we have a routing strategies traditionally build. Based on the parameters collected by virtual agent, routing strategy executes and determines a skilled agent but might not be appropriate everytime. So today, we thought this can be resolved by AI or provide AI prediction as an additional parameter in determining the right Queue and right agent. In this demo, We have defined 5 Queue and given a Name as shown here. Q1_GB_CHAT: Agents belonging to this queue, can help with general banking queries like account information, online bank, branch address, contact number only. And this agent can speak english, spanish. Q1_GB_CHAT_ESP: Agents belonging to this queue, can help with general banking queries like account information, And this agent can speak only spanish. Q2_HL_CHAT_ENG: Agents belonging to this queue can help with mortgage loans, home loans, home loan eligibility. The agent belonging to this queue can only speak english. Q3_CC_CHAT: Agent belonging to this queue can help anything related to credit card. Q4_FR_CHAT: Agent belonging to this queue helps with fraudulent activity on credit card. So we have given simple description or defining this Queue. We have created a Prompt template, with this description & chat summary injected dynamically into Prompt template. This prompt is sent to Text-bison model, which responds with the Queue ID.

PaLMGenerative AI Studio

ReadCast AI - Your Personal AI Podcaster

With the current information overload, many valuable articles are being lost in the endless "read it later" lists. ReadCast AI, an innovative project powered by Google Vertex AI, seeks to transform this paradigm. By using a simple, accessible Telegram bot interface, users can send an online article URL, and ReadCast AI springs into action. ReadCast AI is built on a combination of cutting-edge AI technologies. The core is an advanced language model that crafts an engaging, custom podcast episode for each submitted article. Our AI host, Samantha, not only reads the article content but also creates an entertaining introduction. Samantha introduces each article with such a captivating style that one could easily forget she's an AI and not an expert radio host. What sets ReadCast AI apart is Samantha's understanding of her mission: to make the vast world of information more manageable and engaging for her listeners. Each episode starts with a unique introduction that builds anticipation and excitement without giving away the article content. This makes every episode a thrilling journey of exploration. Moreover, reminding listeners they chose the topic fosters a sense of community and participation. By utilizing Google's Speech-to-Text APIs, we can generate lifelike voice narrations, creating an immersive listening experience. Additionally, our use of Palm2 APIs aids in content parsing and intro creation, ensuring every podcast episode is a unique, information-rich experience. ReadCast AI isn't just a podcast generatorβ€”it's a solution to information overload. It's your personal AI podcaster, ready to turn your neglected "read it later" list into engaging audio content that you can enjoy anytime, anywhere. Welcome to the world of ReadCast AI: Where your saved articles find their voice.


Changing Lives 23 Climate Intelligence Project

At ChangingLives Corp, we understand the challenges that extreme weather events pose to businesses and individuals alike. Traditional weather forecasts often fall short in providing the accuracy and specificity needed to effectively plan for these events, leading to unexpected disruptions and costs. To address this, we have developed an AI-powered Climate Intelligence Platform that provides highly accurate, timely, and personalized predictions of extreme weather events. Our platform leverages advanced data science and machine learning techniques to analyze a vast array of climate data, generating detailed forecasts that go beyond traditional weather predictions. The global weather forecasting services market size was valued at USD 2.2 billion in 2021 and is expected to expand at a compound annual growth rate (CAGR) of 6.7% from 2022 to 2030. The market size is projected to grow from an estimated USD 1.7 billion in 2021 to USD 2.7 billion by 2026, at a CAGR of 9.9% during the forecast period. The weather forecasting services market is set to grow by USD 1,092.95 million from 2022 to 2027, progressing at a CAGR of 9.93% during the forecast period. The market size is projected to reach USD 3.78 Billion by 2030, growing at a CAGR of 6.53%. These figures suggest a significant opportunity in the weather forecasting services market. The growth is likely driven by the increasing need for accurate weather predictions across various sectors, including agriculture, energy, insurance, and more. As extreme weather events become more frequent due to climate change, the demand for advanced weather forecasting services is expected to continue to grow.

ChangingLives23 - Globally
PaLMGenerative AI Studio

ChatBot for Autism Kids

Communication is a fundamental aspect of human interaction, but for children with autism, it can often be a challenge. ChatBot for Autism Kids is designed to enhance communication abilities in children with autism. The ChatBot offers three distinct modes tailored to address the diverse needs of children at different developmental stages. The first mode, augmentative and alternative communication (AAC), provides visual aids such as symbols, pictures, or icons to facilitate understanding and expression. This mode enables children to effectively convey their needs, wants, and thoughts visually, fostering independence and reducing frustration. The second mode is text communication, which allows children who are more comfortable with written language to engage in meaningful conversations. The ChatBot offers a user-friendly interface where children can type out their messages and receive responses, enabling them to express themselves through the written word. The third mode is speech communication, a groundbreaking feature that utilizes advanced speech recognition technology. Through this mode, children can use their own voice to communicate, with the ChatBot understanding and responding accordingly. This not only promotes speech development but also provides a sense of empowerment and self-confidence to children who may struggle with verbal communication. One of the key advantages of our ChatBot is its simplicity and ease of use. It can be easily programmed by speech and occupational therapists, as well as other key stakeholders involved in the child's development. This means that no specialized computer skills are required, ensuring that caregivers and educators can easily customize the ChatBot to meet the specific needs of each child.

application badge
Stable DiffusionPaLMChirp


Our project is centered around designing an innovative Natural Language Processing tailored to process and understand financial data. To achieve this, we're utilizing artificial intelligence technologies provided by Google, and harnessing rich data from the U.S. Securities and Exchange Commission (SEC). We've chosen Google's PaLM 2 as our main NLP model due to its state-of-the-art capabilities in language interpretation. NLP, an AI discipline, empowers computers to comprehend and interact with human language. With its understanding of language nuances, PaLM 2 can interpret and process financial documents, and it can then fetch and rephrase information to the user. At the heart of our project are SEC filings, including forms 10-K, 10-Q, and 8-K. To ensure our system always accesses the most recent data, we're tapping into the SEC's EDGAR system. EDGAR, or Electronic Data Gathering, Analysis, and Retrieval system, is a digital database provided by the SEC. This platform is where all U.S. publicly traded companies are mandated to submit their regulatory filings. These documents are essential sources of financial data. The 10-K form provides a comprehensive annual overview of a company's financial performance. The 10-Q form updates on the company's financial status on a quarterly basis, while the 8-K reports unscheduled material events or corporate changes that could be significant to shareholders or the SEC. Through processing these forms, our system can extract crucial financial details for each company, such as revenue, net income, and cash flow. This extraction is invaluable for investors, analysts, and financial institutions who need the most current, detailed financial information to make informed decisions. In addition, these filings are often hundreds of pages long, so an LLM that integrates directly into the source of the filing can efficiently fetch relevant information for our users.


MLSquad - Special Victims Mentor

This web app employs advanced AI techniques to analyze and interpret medical forensic reports of sexual assault crimes in order to generate a detailed timeline of events and calculate the remaining time under the Statute of Limitations for potential prosecution. Key Features: PDF Upload and OCR: Users can upload PDFs of medical forensic reports. The app uses OCR (Optical Character Recognition) technology to convert the images and scanned texts into machine-readable format. Report Parsing: A Large Language Model, like GPT-4, is used to parse and understand the content of the reports. It identifies key dates, incidents, and entities mentioned in the text. Timeline Generation: Based on the parsed data, the app generates a visual timeline of events related to the assault case. It highlights significant milestones such as the date of the crime, date of the medical exam, when the crime was reported to authorities, etc. Statute of Limitations Calculator: The app also calculates the remaining time under the Statute of Limitations for the particular crime in the jurisdiction it occurred. It uses a database of legal information (updated regularly) to determine the specific Statute of Limitations for sexual assault crimes in different jurisdictions. Updates and Notifications: The app keeps track of the dates and sends reminders to the concerned parties (like the victim, lawyers, or law enforcement authorities) about the approaching end of the Statute of Limitations. Security and Confidentiality: Given the sensitive nature of the information, the app ensures robust security and privacy measures. Data is encrypted at rest and in transit, and strict access controls are in place. The app complies with all relevant legal and ethical guidelines, including HIPAA.

ML Squad