Cohere
Cohere API provides access to state of the art, affordable, and easy-to-deploy large language models capable of powering the next generation of game-changing AI native applications.
Powerful NLP Model
- Cohere trained their models on billions of words to understand the nuances and context markers of human communication.
Flexible API
- The Cohere API works with many different libraries that fit every stack. Cohere's Python, Node, and Go SDKs make AI easy to integrate into your app.
The Future of software
- Build innovative and extremely powerful applications that were previously not possible. The future of software development is AI native.
Cohere Challenge
👉 The challenge for this hackathon is to build the most creative and innovative Semantic Search application using Cohere API.
Prizes and Awards
🥇 $5000 API credits + $1000 cash
🏆 $4000 API credits + $1300 cash to be distributed among the finalists
Additionally, all winning teams will receive exclusive Cohere Swag and have the chance to: Meet Cohere Founder for a virtual coffee & record a video of their demo that will be promoted on Cohere's channels!

What is Semantic Search?
Language models give computers the ability to search by meaning and go beyond searching by matching keywords. This capability is called semantic search.
⚡ Powering a Private Search Engine
Popular use case of semantic search is building a next generation web search engine. Impressive, but the applications of semantic search go beyond that! They can empower a private search engine for internal documents or records. They can be used to power features like StackOverflow's "similar questions" feature. And you can build many more things with it.
🖹 Semantic Search and Text Sources
Semantic search is the most successful with text sources where the answer to a query is likely to be in a single, concrete paragraph, such as technical documentation or wikis which are organized as a list of instructions or facts.
🏗️ Build a State of the Art Application
Get the creative juices flowing and build a state of the art application of semantic search we haven’t seen yet!
Semantic Search Sandbox + Resources
We encourage you to explore semantic search with Basic Semantic Search notebook, Cohere’s docs and Toy Semantic Search sandbox. Sandbox is a collection of experimental, open-source GitHub repositories by Cohere that make building applications for developers fast and easy, regardless of ML experience.
👉 Check out basic Semantic Search example notebook by Cohere
👉 Build a simple Semantic Search engine with Cohere
👉 Check out docs on how to build a Try out semantic search
👉 Try out Toy Semantic Search sandbox
Multilingual Semantic Search
Text embeddings are a central component in machine language understanding. They are numeric representations of text (be it a document, an email, or even a sentence). An embedding model translates text into a list of numbers that capture its meaning. A multilingual embedding model is able to do that well for many languages.
This video demonstrates Cohere's multilingual embedding model, and its ability to represent many languages.
👉 Check out multilingual model Github repo
👉 Try out this cool multilingual movies search and recommendation demo

API Access
Signup for API Access
Cohere's API is currently free-to-use for everyone. Sign up for Cohere and start integrating NLP into your builds now!
Semantic Search Hackathon details
Join lablab and Cohere for a week to innovate and build the new generation of NLP powered applications. Find all the relevant details below.
🗓️ Where and when
The hackathon starts on December 16th and ends on December 23rd. Over the weekend, you'll have the opportunity to learn from Cohere and lablab experts during workshops, keynotes, and mentoring sessions. The hackathon will take place on the lablab.ai platform.
🦸🏼♂️ Who should participate?
Previous experience in AI is not required but welcomed. While many participants are industry experts, we also welcome people with other types of domain knowledge that want to understand & explore how AI can be used in their fields.
🔐 Access to Cohere API
To get started with Cohere NLP API please signup using the following link: https://cohere.ai/signup. Your trial API key are free and has and can handle up to 100 calls per minute free of charge. You can find more information about the API here.
😅 How about teams?
If you don’t have a team you will be able to match and team up with other participants around the world. Finding & creating teams can be done from the dashboard you can access after you enroll. We also recommend checking our Discord server to find teammates and discuss ideas. You can join it here
🛠️ How to participate in the hackathon
The hackathon will take place online on lablab.ai platform and lablab.ai Discord Server. Please register for both in order to participate. To participate click the "Enroll" button at the bottom of the page and read our Hackathon Guidelines.
🧠 Get prepared
To get prepared for the hackathon, we recommend you to start at our Cohere technology page where you can find all the relevant information about the API and how to use it plus cohere tutorials and cohere boilerplates.
Applications build on Cohere
Learn about the winning projects from previous episodes of the Cohere hackathons.
Speakers, Mentors and Organizers
Hackathon FAQ
Who can join the Hackathon?
We welcome domain experts from all industries, not just AI or tech. Successful AI solutions require a combination of technical expertise and domain knowledge. Coding experience is recommended.
Do I need a team?
You are welcome to join as a team or solo, if solo. We encourage you to look for a team before the event. We recommend you to join the Deep Learning Labs Discord channel: https://discord.gg/gCuBwBB35k and posting in the #looking-for-team channel to get to know your potential future team members.
Do I need a Github account?
It is recommended, that at least one team member has a Github account. You can create one for free if you don't already have one.
I have other questions.
Feel free to reach us on social media, or through our Discord channel.
Event Schedule
- To be announced
Winner Submissions 🏆
.png&w=3840&q=75)
Found In Translation
Our solution is a smart Slack bot that implements multilingual semantic search and sentiment analysis to facilitate multi-language messaging platforms by improving the chat search function. The bot allows users to search using any language for messages in any language. It also contains a sentiment analysis that gives moderators a sentiment report on the messages of specific users or the chat as a whole.
RatLabs

Summit
Summit helps you search and summarise research papers to quickly get an overview of what the research says, saving you time and effort in your quest for knowledge.
Grounded AI

SemantoTube
Our application aims to make it easier for users to find and access the information they need within the specified video. By using the Cohere API, we are able to perform advanced semantic searches that go beyond simple keyword matching, allowing users to find content relevant to their query in the specified video. Our application utilizes the Cohere API to locate and highlight the specific transcript within the video that matches their search. This makes it easier for users to quickly find and access the information they need, without having to manually search through the entire video. Overall, our application is a game-changer for anyone looking to find and access specific information within YouTube videos. Whether you're a student looking for a specific lecture, a researcher trying to locate a specific quote, or just someone looking for information on a particular topic, our application makes it easier than ever to find what you need.
Giga Chads
Submitted concepts, prototypes and pitches
Submissions from the teams participating in the Semantic Search AI Hackathon event and making it to the end 👊
.png&w=3840&q=75)
Found In Translation
Our solution is a smart Slack bot that implements multilingual semantic search and sentiment analysis to facilitate multi-language messaging platforms by improving the chat search function. The bot allows users to search using any language for messages in any language. It also contains a sentiment analysis that gives moderators a sentiment report on the messages of specific users or the chat as a whole.
RatLabs

Summit
Summit helps you search and summarise research papers to quickly get an overview of what the research says, saving you time and effort in your quest for knowledge.
Grounded AI

SemantoTube
Our application aims to make it easier for users to find and access the information they need within the specified video. By using the Cohere API, we are able to perform advanced semantic searches that go beyond simple keyword matching, allowing users to find content relevant to their query in the specified video. Our application utilizes the Cohere API to locate and highlight the specific transcript within the video that matches their search. This makes it easier for users to quickly find and access the information they need, without having to manually search through the entire video. Overall, our application is a game-changer for anyone looking to find and access specific information within YouTube videos. Whether you're a student looking for a specific lecture, a researcher trying to locate a specific quote, or just someone looking for information on a particular topic, our application makes it easier than ever to find what you need.
Giga Chads
.png&w=3840&q=75)
AlphaBot
Around 42% of start-ups fail due to a lack of market-related knowledge. Our tool helps start-ups or small businesses or even people with great ideas, to understand the market, the market need, and the financial aspect of the industry that they want to venture in.
AlphaBot

Emoji Character Voice Recog
A voice recognition way of creating emojis and special characters. 10 emojis and 13 special characters are initialized. When we speak with initialized key(there names in code), they will be displayed.
Sky High Tech

Semantic search models
Semantic search has been used in daily life applications, especially for the elderly people in speech-recognition, because they are not able to use touch device due to their bad eyesight. Speech recognition is one of the applications of semantic search, and for this there have been used NLP and Preprocessing. As a result, the information is gathered as a cluster, which then embedded into a system. Moreover, the politics has also been analyzed through preprocessing which is branch of NLP that is being used in semantic search.
AI isnt AI

Libris
Libris is a tool for making knowledge accessible to all. You just describe an important topic that you want to investigate and Libris will bring to light the relevant excerpts from the most brilliant minds of all time. You can also filter by specific authors and titles to search in a specific subset of the bibliography that is specially relevant to your research. The magical thing about Libris is that it doesn't matter the language you speak or the language the text was written on, the experience is seamless: you write the concept in your own language and get results in your own language. You can even access the original work for further reference!
Going Solo

ezInvest
ezInvest is an application, that helps user in making a decision to get a Stock based on News headlines Sentiments. Investment in Stocks is very Risky. Many Cases like Political, Inflation, economics, investors behavior Impacts Stock Prices. Identifying Stocks to Invest is Challenging. Existing Applications are very Expensive. ezInvest will address all these issues.
Team Deepfai

LinguaScope
LinguaScope helps refugees query official immigation documentation in their own language for urgent questions when they seek asylum. Users can query for childcare, getting heathcare etc. Often it is not possible to get translators and this app will help them navigate resources for critical information.
LinguaScope

News Media Monitoring
News monitoring news allow observing and collecting key customer data directly from mentions and discussions about your company (or your client’s company). These discussions take place in social media comments, product review websites, blogs, podcasts, discussion forums, and media websites.
Embeddings

whereDoYouMean
When students are revising their work, they face a few hurdles - (1) finding content from a big repository of school materials is time-consuming, especially when the student does not know specifically where to look for the information; (2) results searched online might not be exactly the same as what the school is teaching in terms of scope. Hence, we decided to create an application where teachers can transcribe their videos and generate the embeddings. Users can then ask a question and the top few links to the relevant videos will be given to the student (with the time-stamp attached). The context is also provided to the student so that they can make a judgement as to which video is worth exploring. After which, students can also vote on the relevance of the videos, which will be used to affect future training. We hope that this project can help students learn better with their school's resources.
whereDoYouMean
.png&w=3840&q=75)
CoHotel
Recommend and book the best matching hotel for travelers to the US based on Semantically Searching Hotel Reviews in the city they are traveling to. Integrates with MyCroft AI to provide text and voice conversational interfaces with feedback loops that increase the accuracy of choice.
CoHotel

CoDoctor
CoDoctor is a healthcare assistance tool that helps users find answers to their questions about different health conditions. It uses a multilingual model called Co:here, which uses a semantic search algorithm to search through a dataset and provide accurate and relevant information based on users query. CoDoctor is a useful resource for anyone looking for reliable information about various health conditions and how to manage them. Whether you have a specific question or just want to learn more about different health conditions in general, CoDoctor can help you find the information you need.
OGNG
.png&w=3840&q=75)
Resume AI
Generate personalized resume customized to the job description from a master resume.
Resume AI

Product Description Generator
This web app let the user to fill in the name of product and its properties as keywords. Then, uses cohere-ai (via Node) to generate the description of the product related to the given keywords. Can be used for advertisement, Digital Marketing,& SEO related purposes.
Non Zero
.png&w=3840&q=75)
Patent Search and Generation
This is a service assists patent attorneys/professionals/applicants/to trawl through patent databases and return the most semantically relevant patents instead of simply returning patents that correspond to lexical/keywords matching AND generate a draft patent application for the inventor. As such, we are able to reduce the time taken for the patent application process,resulting in greater revenues for patent attorneys/professionals while extending the useful patent life ofthe inventor's patent, as well as expedite his patent application process
Love AP

FilmyCode
We made an app to search movies. We have used cohere and pinecone. Pinecone is being used for vector database and cohere is for embeddings. Our app gets movies based on user inputted query. So if a user search "Alien invasion movie", the app outputs "Edge of tomorrow, etc". It is mostly google like search but for movies and also we are using NLP (cohere-large) model.
Sentient

Project Eval
Eval aims to address the problem of subjectively evaluating test answers. Traditionally, this task has been carried out manually by human graders, which can be time-consuming and prone to bias. To address this issue, the project utilizes Cohere powered APIs to automate the evaluation process. The use of Cohere APIs allows for the integration of advanced natural language processing techniques, enabling the system to accurately understand and analyze the content of test answers. The custom model built upon these APIs then scores the answers based on suitable metrics, which can be tailored to the specific requirements of the test or assessment. One potential application of this technology is in the field of education, where it could be used to grade assignments or exams in a more efficient and unbiased manner. It could also be utilized in professional settings for evaluating job applications or performance evaluations. In addition to increasing efficiency and reducing bias, the use of automated evaluation techniques has the potential to provide more consistent and reliable scoring. This can help to ensure that test-takers receive fair and accurate assessments of their knowledge and skills. The model for the same was evaluated based on 4 major metrics: - Semantic Search: this is the primary scoring strategy of Eval. It is used to semantically understand the answer given and evaluate based on content rather than simply scoring based on textual similarities. Cohere Embed was used to generate embeddings for 5 suggested answers for the question and the answer to be checked. Then we find the distance from the nearest neighbor out of the 5 suggestions and the answer. This distance is used to grade the answer. - Duplication Check: partially correct answers with duplication of text tended to get higher similarity scores compared to the ones without duplication. To stop students from using this exploit to gain extra marks, a duplication checker was implemented based on Jaccard-Similarity between sentences within the answer. - Grammar Check: this strategy aims to check the grammar of the answer and assign a score based on the number of grammatical errors. We used Cohere Generate endpoint to generate a grammatically correct version of the answer, then check for cosine similarity of the generated version with original version to check if the original version was grammatically correct. - Toxicity Check: this aims to detect for toxic content in the answer and penalize an answer if it is toxic. We trained a custom classification model on Cohere using the Social Media Toxicity Dataset by SurgeAI which gave a 98% precision on the test split. We also implemented a Custom Checks which allows users to give different weights to each of the three different metrics based on how important they are for the evaluation of the answer. This allows for a more personalized evaluation of the answer. We built our custom model into a Flask-based REST API server deployed on Replit to streamline usage and allow people to access the full-functionality of the model. We also built a highly interactive UI that allows for users to easily interact with the API and evaluate their answers as well as submit questions.
chAI
.png&w=3840&q=75)
Genesis
It is an search app to look up meaning, hope, faith, etc in the Bible based on the user situation (happiness, sadness, depression, motivation, ...)
Genesis

Profile Picker
For startups and new businesses, it's hard to get budget, manpower and time to put on resume sorting kind of works. When large number of applicants apply for a job role then it's really difficult to choose the best ones by going through each resume. Our product "Profile Picker" provides businesses ability to sort best resumes based on the job description. Our ML model understands the meaning of the words that are in the resumes and in the job descriptions and it also gives multilingual support which means job description and resumes can be written in different languages.
VigyaAI

ZenWork
ZenWork is an MVP semantic search engine for finding your dream company that shares your vision.
Sepik

Cofinder
Built by the Community, for the Community A semantic search tool enabling Cohere users to find relevant content in one place based on their personal goal. The aim is to breakdown barriers for users such as developers, entrepreneurs and Data Scientists, making it easier to find the information they need, to remove the effort of obtaining information and focus on building amazing applications using Cohere's language models.
Cofinder

Cohire
Cohire is an AI powered job portal that allows top employers to find the best talents to hire. Contrary to AI taking over our jobs, this project tries to take a different narrative where AI will help us find jobs, not replace us.
Pandora

sundarEduAI
sundarEduAI is an advanced AI learning tool which helps you to find answers of questions of given topics/chapters within few seconds. It also increase accessibility for users who want find solutions quickly. It's quite fast, easy-to-use & accurate. It'll be next-generation of edtech.
sundarAI

AI Generation
In my Project you can get blogpost content for posting you blogs and you can have the trending hashtags and you can also generate mails
Alpha

Embetter support for Cohere
By supporting Cohere via embetter, we gain scikit-learn compatibility. That means it's easy to: 1. Build active learning pipelines via partial_fit 2. Use it for bulk labelling via bulk 3. Use it to find bad labels via doubtlab 4. Run quick retreival projects via simsity And more!
calmcode

Indic News search engine
🌍 The Solution uses a multilingual semantic model from COhere to 🚀 revolutionize the media and news industry in multilingual markets like India(We have used telugu , hindi ,bengali,English news dataset for this use case), allowing anyone to track 📰 regional news in real-time without the need for translation or understanding of other regional languages. 🙌
team phoeniks

Citation needed
Citation Needed is a fact verification assistant that allows you to find relevant sources and citations to a given claim. Specifically, we begin with a information database, where each claim or fact is substantiated with a reference. From there, a query claim can be used with semantic search to find relevant sentences, and hence the citations of them. This application can be useful to a variety of audiences, from researchers looking to more efficiently browse through papers, teachers looking to better fact check essays, to even students looking to find more reading material. We believe that Citation Needed can be a powerful tool if scaled up fully. The current demo has a "Narrow" mode, which only contains topics related to neural networks, and a "General" mode, which scrapes for additional information on the fly, after extracting important subjects through the use of the cohere text generation API.
Citation Needed
.png&w=3840&q=75)
Search Light
A semantic search system that enables you to search your codebase using natural language. Project repo https://github.com/silvererudite/code_search
Amadeus

Flash Learning
Flash Learning is a flashcard application that helps the user to retain any material, like terms, definitions, laws, etc, through active recall. Unlike the classic flashcards, where the process of evaluation is ambiguous because the users say their answers verbally, our application offers the opportunity to get AI-based self-assessment scores that quantify their level of knowledge of the studied card. These scores can serve as indicators on how well the users know the material and if they are ready to learn something else or end their learning session. It can be used for both formal and informal education, where the instructor can prepare public flashcards with the curriculum needed for tests and exams. The Leitner System - a widely used method of efficiently using flashcards - is now implemented digitally in our application. It is a simple implementation of the principle of spaced repetition, where cards are reviewed at increasing intervals to enhance the process of learning and store the information in the long-term memory. The users can choose to know the card at the GOOD (>80%) or EXCELLENT (>90%) level based on their necessity. If their score is not excellent, they will even get explanations with GenerativeAI to help them improve their score.
FlashLearners

Quick Learners
Our education platform aims to revolutionize the way students learn by providing a unique and personalized learning experience. Through the use of semantic search, we enable users to easily find specific moments in a video that discuss a particular concept, making it easier for them to focus on and understand the material. Additionally, we recommend people with similar interests to connect with each other, fostering a sense of community and collaboration among learners. Lastly, we recommend posts to users based on their interests, ensuring that they are constantly exposed to relevant and engaging content of the user with same interests.
Breaking AI
.png&w=3840&q=75)
Podsearch
Podcasts are an excellent source of knowledge. But they can be too long and hard to pay attention to it the entire time. What if there is a more intuitive way to search for podcasts and also for info within podcasts? This is where our product comes into play. Key highlights 1. Searching for podcasts suited to your taste 2. Searching for answers within a podcast itself by asking it queries and without listening 3. Marking exactly where the answer is and summarising it. 4. Telling user what queries this podcast answers Major Uplifts: 1. Generating queries for dialogues in transcript using the prompt - "Generate 5 questions for the following passage {passage}" 2. Training a classifier using cohere api using the generated queries and dialogues 3. Highly scalable architecture 4. Podcast is just an example. Most documentation (python libraries, eth doc) have only keyword search. It is possible to scrape the documentation and build an index for a search engine using our architecture easily.
Info Insighters

Coherent Speech
Multilingual voice content generator, powered by co;her(💜+🤖) Generate model. Using the power of whiper, co;her(💜+🤖) and google to create speech content generator. This Application helps in generating audio content using speech as a prompt.
OraOraOra
.png&w=3840&q=75)
YTBrief
AI solution that will make your YouTube experience easier by getting answers from videos in seconds.
TopGeez
Teams: Semantic Search AI Hackathon
Check out the roster and find teams to join