VectorShift

I have a big announcement! Today, we are launching VectorShift’s AI Assistant product. VectorShift is a No-Code Generative AI platform. We allow you to build AI workflows automations without needing to code or understand AI! With the AI assistant product, directly place an assistant interface on your AI workflow and export to end-users via URL, iFrame, API, or as a WhatsAPP / SMS / Slack bot.

VectorShift (YC S23) is a no-code generative AI platform. With it, you can search through knowledge bases and deploy workflow automations, all while live-syncing with your data and tools. Today, they're launching their new chat/assistant product: directly place a chat assistant interface in your AI workflow and deploy it internally within your organization or on your website or product. Alex Leonardi and Albert Mao both studied computer science and statistics together at Harvard. While in college, they grew a business to 7 figures in revenue. 6 months ago, they left their jobs at Blackstone and McKinsey— where they deployed AI/ML solutions for portfolio companies and helped Fortune 500 organizations digitally transform with revamped GTM motions, respectively— to found Vectorshift. With VectorShift, they're helping teams leverage AI without needing to code or AI knowledge. Learn more at https://lnkd.in/gQqYMwZF. Congrats on the launch, Alex and Albert!

Since starting up, I have learned that everything is in execution. Today, we 🚀launch 🚀 one of the largest VectorShift product updates yet! See below what we cooked up just in the last few weeks: 1. Version Control: Revert back to a previous deployed version of a pipeline. This is helpful if you made changes to a pipeline that you do not want. 2. Auto-Saving: pipelines will auto-save - never lose your progress again! 3. Multiplayer Mode: work on a pipeline together with teammates and see changes updated live (e.g., new nodes, new connections). Automatically enabled for all accounts. 4. Offline Mode: Continue to edit pipelines if you lose internet connection. When your computer gets access to internet again, the pipeline will save. 5. Search Interface: place a search interface on a pipeline or knowledge base. Do so within the pipeline builder (Deploy >> Search) or on the new “Search” tab. 6. Enhanced CSV reader: we made significant accuracy improvements in the ability to query over tabular data. Utilize the CSV query node under the “Data Loader “ tab. 7. Dropbox: live-sync with your dropbox and leverage the data in your AI workflows. 8. One Drive: live-sync with your OneDrive and leverage the data in your AI workflows. 9. S3: live-sync with your S3 storage and leverage the data in your AI workflows. 10. Confluence: live-sync with your Confluence knowledge base and leverage the data in your AI workflows. 11. Slack (private channels): in addition to public channels, live-sync with data from Slack private channels and leverage the data in your AI workflows. 12. Zapier / Make nodes: have a pipeline run a Zapier / Make.com workflow. 13. Zendesk: search tickets, create new tickets, and trigger a pipeline with a new Zendesk ticket. 14. Notion (create new block): add text to an existing Notion page. 15. Postgres: connect and query your Postgres database. 16. Pinecone: connect and query your Pinecone database. 17. Re-scrape URLs: when loading URL data into a knowledge base, define how often you want the data to be re-scraped and re-embedded. This will ensure the data is fresh. 18. Citations: Enable LLM to site sources in the output (URLs and files). Access by clicking on the gear on an LLM. 19. Significant improvements to chatbots: Overall UI improvements, edit display details, add assistant / launcher / avatar image, enable AI generated follow up questions, copy and download responses 20. Map Node: map a function to transform input data to output data. You can find this node under the “Logic'“ section of the No-Code builder. 21. LLM function calling: provide LLM access to tools (LLMs will have a handle called “Tools”). If the LLM deems it is necessary, it will leverage the tool in product output. We have built this with full flexibility - you can connect anything to the tool handle (e.g., web search, calling an API, etc.). 22. New Models: Anthropic Claude 3, Gemini 1.5 pro, Cohere Command-r and Command-r plus

We automated all our customer support emails at VectorShift! Majority of the customer support emails I get can be answered through our documentation, videos, FAQs, or meeting transcripts from support calls. We build a pipeline that searches and live-syncs to a knowledge base with our website, google drive, and other tools to automatically draft email responses when emails come in!

March 5 - VectorShift Update Use Case Highlight: With VectorShift, you can place triggers across your applications (e.g., new typeform message, new slack message, new email), create AI workflows with our no-code builder, and perform actions (e.g., send a slack message, send an email, update a lead in CRM).For example, we have seen many users use their customer support / lead collection Typeform as a trigger for sending a slack message (to help coordinate who to take action) or to send a personalized email to the user (e.g., to schedule a demo). Product Updates In the two weeks, we shipped: 1. Start from template form: clicking “new” on the pipeline page will bring up a page where you can either start from a blank canvas or from a no-code template. Hopefully this will make it easier to get started building! 2. New Data Sources / Integrations: Github (new commit trigger, new PR trigger, new issue trigger, search files, read files, create a PR, update PR), Notion (create a new page), Hubspot (new contact trigger, new company trigger, new deal trigger, new ticket trigger, new engagement - notes trigger, new engagement - tasks trigger, new engagement - calls trigger, search items, fetch contacts, fetch companies, fetch deals, fetch notes, fetch calls) 3. Deploy to Slack App: deploy a chatbot directly as a slack app. Can be found in the “Export” tab under chatbots. 4. Cron Job: schedule your workflows to deploy at predetermined time intervals (e.g., every hour, day, week, month). 5. Gemini Vision Model: Google’s new Gemini multi-modal model is now on our platform! 6. Variable Creation: within LLM and text nodes, you can insert a variable by clicking the insert variable button. Variables will also be displayed differently than text as well.

I’m super excited to share that VectorShift has raised $3m seed round 🎉. Thank you to Y Combinator (Harj Taggar, Richard Aberman), 1984 Ventures (Ramy Adeeb, Mark Percival, Samit Kalra), Defy VC (Bob Rosin, Neil Sequeira, Brian Rothenberg, Adam P.),  Formus Capital (Cem Garih, Fethi Sabancı Kamışlı, Burak Karahasan, Emir Bölen, CFA), Soma Capital (Nikhita Jaaswal), 468 Ventures (Bernhard Janke, Bardo Droege),  Pioneer VC, and many more incredible investors on our cap table. 6 months ago, I was working at McKinsey & Company. I would stay up every night to 2-3am burning the midnight oil churning out excel models and powerpoint slides. Despite this, McKinsey taught me how to present to clients, engage with stakeholders, and approach every problem with analytical rigor. We applied to Y Combinator on a whim. Despite the fact that I am relatively non-technical, our group partners, Harj Taggar and Richard Aberman welcomed me with open arms and introduced me to the startup world. Thank you for your generosity, support, and advice. I couldn’t have imagined the journey since. Going through the Y Combinator S23 batch, I met some of the most humble and innovative people tackling the world's largest challenges. Founders who truly want to make a difference and change the world. Thank you to you all for being a part of my life and appreciate the opportunity to be a part of a startup community. At VectorShift, we are building the generative AI platform that allows businesses to build custom AI solutions for their repetitive tasks. We’ve helped thousands of teams already build AI search engines, assistants, and workflow automations. During my time at McKinsey, I met countless people from organizations of all sizes doing manual tasks and we want to make going to work better for you. With VectorShift, you can  + Centralize and live-sync your data across your applications + Search across your knowledge base + Define triggers and perform actions across your tools  + Build AI workflows either through no-code or through python SDK with complete interoperability Finally, I would like to thank the amazing VectorShift Team and my co-founder Alex Leonardi. We have the hardest working team - and we ship feature after feature after feature. I’m super excited for the future of VectorShift. Hard work consistently trumps other factors. If you are looking to leverage generative AI in your company, feel free to reach out!

Automate your customer support between typeform and slack with AI! New typeform messages will: 1. Trigger a VectorShift AI workflow 2. Performs actions throughout your toolset (summarize the request with a slack message, send a personalized email) #ai #ml #genai

How do you reduce LLM hallucinations? I spent a good portion of every week helping VectorShift customers reduce AI hallucinations in their workflow automations so I put together this guide that discusses a few helpful techniques! #ai #genai

With a few clicks, you can place a ChatGPT interface on a VectorShift AI workflow! Great way to share internally or to deliver an AI solution for clients.

Tons of new VectorShift product features from LLM evaluations product, new LLM models, and tons of new automations (triggers and actions) ___________ February 14 - VectorShift Weekly Update Weekly Use Case Search and live-sync your data across tools With VectorShift, you are able to live-sync and embed your data from Google Drive, OneDrive, Notion, Airtable, Slack, and more. Employ semantic search across all your data to get answers and boost your productivity! Product Updates In the past week, we shipped: 1. Evaluations: test pipelines across test cases and create pipeline “scenarios” where you can adjust prompts, architectures, parameters, and more while viewing changes in output. 2. Custom API Node: make an API request to any REST endpoint. You can find this node under the data loaders tab. 3. Enhanced web search: we added capability to use you.com and Exa AI to query the web. You can find these nodes under data loaders. 4. GPT 4 vision: convert images to text! 5. Automations: Google calendar trigger (new event and updated event), Google calendar action (new event), Google Drive trigger (new file, new folder, and updated file), Google Docs action (append text to document), and Copper 6. CRM action (write new leads). 7. Date Node: pass the current time to vector stores or LLMs. 8. New embedding models: customize which embedding model you use in the vector query node! Podcast Feature: Tim Qin, CEO of Roame: Business class flights US to Japan for less than $1,000 Business class flights US to Japan for less than $1,000 - Roame's CEO Tim Qin New LLM Guides: Guide to prompt engineering: This guide summarizes key prompt engineering strategies and offers practical explanations and prompt examples, helping increase the efficiency and accuracy of LLM output. How to get better output with Anthropic’s Claude: This article explains the specifics of Claude LLM and offers practical tips and tricks for designing prompting techniques when working with Anthropic's latest Claude AI chatbots.