DanswerAI

DanswerAI

Technology, Information and Media

Ask questions in natural language and get back answers backed by private sources such as Slack, GitHub, Confluence etc.

About us

OpenSource Question-Answering System. Ask questions in natural language and get back answers backed by private sources such as Slack, GitHub, Confluence, etc.

Industry
Technology, Information and Media
Company size
2-10 employees
Type
Privately Held
Founded
2023
Specialties
Information Retrieval, NLP, and Generative AI

Employees at DanswerAI

Updates

  • View organization page for DanswerAI, graphic

    1,399 followers

    The Danswer team just leveled up AGAIN with Richard Kuo joining as our second Founding Engineer! Richard brings rich experience from his remarkable journey scaling up Youtube in their early days as well as being a pillar of engineering and product for industry giants like Google and Netflix. Not only that, Richard has a flame for entrepreneurship - he was even a founder and operator of his own startup previously! In his first few days Richard has rearchitected the most complicated flows of the Danswer backend, replacing a convoluted locking mechanism with a much more graceful queue based implementation that resolved a number of perplexing deadlock situations. We are SOO thrilled to have Richard as a teammate and so excited for what we’ll all build together

    • No alternative text description for this image
  • View organization page for DanswerAI, graphic

    1,399 followers

    Check out how AutoScout24 uses Danswer to introduce GenAI into their processes and help their support team be more efficient.

    View profile for Arslan Mehboob, graphic

    Platform Engineer @ AutoScout24 | AWS Certified Solutions Architect | AI Enthusiast

    I got the privilege of writing a blog post in AutoScout24's tech blog. In it, I dive into how we leveraged AI (LLM) to enhance our internal support system. What began as a simple Hackaday project has now evolved into a crucial part of our support process. We utilized DanswerAI to integrate and connect different components, with AWS Bedrock powering the underlying LLM. Check it out and let me know what you think! https://lnkd.in/dnX2Tacs

    A tale of testing AI in the wild - Our experience with automating internal platform support with a LLM bot

    A tale of testing AI in the wild - Our experience with automating internal platform support with a LLM bot

    tech.autoscout24.com

  • View organization page for DanswerAI, graphic

    1,399 followers

    View profile for Chris Weaver, graphic

    Co-Founder @ Danswer (YC W24)

    Here’s how a needle in a haystack helps us build better LLMs: Say you have a model that can ingest 1m tokens as input (e.g. Gemini 1.5 Pro). You might wonder whether it could provide precise answers given such a large body of text. Enter the Needle In A Haystack test: 1. We take all the text from the first 5 Harry Potter books (roughly 0.9m tokens). 2. We pick a random place in the body of text and add a phrase like “The super secret gemini code is danswer.” 3. Then we input all of the text into the model prompt and append the question: “What is the super secret gemini code?”. 4. If it’s a good model it should answer with “danswer”.

    • No alternative text description for this image
  • View organization page for DanswerAI, graphic

    1,399 followers

    View profile for Chris Weaver, graphic

    Co-Founder @ Danswer (YC W24)

    It’s been a busy month but we keep shipping! Here are some of the enterprise-ready features we released this month: ⭐️ New Connectors (Salesforce, Teams, Dropbox, Discourse, and more!) 👉 Fast AI-Model Switching (Use your favorite AI model) 🔧 AI Tool Calling + Custom Tools 👀 Web Search Capability 🖼️ Image Generation 📨 Share Chats with Your Team Shout out to Danswer’s open-source contributors for helping us integrate Microsoft Teams, Dropbox, and Salesforce!

    • No alternative text description for this image
  • DanswerAI reposted this

    View profile for Chris Weaver, graphic

    Co-Founder @ Danswer (YC W24)

    We've noticed that all of our enterprise customers at Danswer have the same favorite feature: Our AI model-agnostic platform. Since we don’t want our enterprise customers to be locked in to a specific vendor, we enable them to use any AI model under the hood—from OpenAI to Claude to Llama—which makes it really easy to switch. This is essential for security and privacy because it means that companies can run Danswer with an open-source AI model on their own hardware—without sharing any data with a third party.

  • View organization page for DanswerAI, graphic

    1,399 followers

    Danswer just got more studious with Pablo Hansen joining as our Founding Engineer! Pablo brings a wealth of academic knowledge on Natural Language Processing and Information Retrieval.  (Fun fact: he graduated with a master’s degree in Artificial Intelligence before his 19th birthday). We’re excited to see how his background will help us improve Danswer’s search performance and redefine our user experience. Pablo’s fastidiousness, attention to detail, and dedication have already made him an integral part of our team, and we look forward to many more sprints together!

    • No alternative text description for this image
  • View organization page for DanswerAI, graphic

    1,399 followers

    Really excited to be a part of Y Combinator! It's been a blast working with all our amazing customers so far, and we're only just beginning 💃 If you're looking for something to make it incredibly easy to find any piece of workplace information without compromising security or data privacy, then get in touch. We'd love to chat!

    View organization page for Y Combinator, graphic

    1,040,234 followers

    DanswerAI (YC W24) is a team-specific ChatGPT that connects to knowledge from workplace tools like Google Drive, Slack, Jira, etc. The best part: it is open source and teams can self-host it for complete control over their data. By making information instantly accessible, Danswer is already helping over 50 organizations to: - Speed up customer support and escalation turnaround. - Enable sales teams to get fuller context and improve efficiency in preparation for calls. - Improve engineering efficiency by making documentation and code changelogs easy to find. - Track customer requests and priorities for product teams. - Self-serve IT, onboarding, HR, etc. Founders Chris Weaver and Yuhong Sun are engineers with intimate knowledge of deep learning, transformers, and information retrieval. While at Robinhood and Alation respectively, they experienced firsthand the downstream effects of difficult-to-find information and are now leveraging LLMs to solve this pressing problem. Learn more at https://lnkd.in/dSgyDK7e. Congrats Chris and Yuhong on the launch — and welcome to YC!

Similar pages

Funding

DanswerAI 1 total round

Last Round

Pre seed

US$ 500.0K

See more info on crunchbase