- 29
- 2 544
Andre Moura
เข้าร่วมเมื่อ 21 มี.ค. 2024
PostHog-LLM Chat analytics: Analyzing Emojis Dialogs | HogQL
I'll dive into how to use HogQL expressions in PostHog to filter for dialogs containing emojis only, analyze trends over time, and breakdown emoji usage by device type and user toxicity.
Regex used: [\\x{1F600}-\\x{1F64F}\\x{1F300}-\\x{1F5FF}\\x{1F680}-\\x{1F6FF}\\x{1F1E6}-\\x{1F1FF}\\x{2600}-\\x{26FF}\\x{2700}-\\x{27BF}\\x{1F900}-\\x{1F9FF}\\x{1FA70}-\\x{1FAFF}]. This is escaped
Project: minuva.com/PostHog-LLM
GitHub Page: github.com/postlang/posthog-llm
Dataset: huggingface.co/datasets/allenai/WildChat-1M
Discord: discord.com/invite/uNurWjDkke
#llm #analytics #monitoring #conversationalai #datainsights #llmops #trends #timeseriesanalysis #insights #conversationalai #selfhosted #userinsights #Chatbot #llms #regex #filters
chatbot analytics llm analysis llm monitoring llm observability Conversational Analytics user analytics llm domain conversations
Regex used: [\\x{1F600}-\\x{1F64F}\\x{1F300}-\\x{1F5FF}\\x{1F680}-\\x{1F6FF}\\x{1F1E6}-\\x{1F1FF}\\x{2600}-\\x{26FF}\\x{2700}-\\x{27BF}\\x{1F900}-\\x{1F9FF}\\x{1FA70}-\\x{1FAFF}]. This is escaped
Project: minuva.com/PostHog-LLM
GitHub Page: github.com/postlang/posthog-llm
Dataset: huggingface.co/datasets/allenai/WildChat-1M
Discord: discord.com/invite/uNurWjDkke
#llm #analytics #monitoring #conversationalai #datainsights #llmops #trends #timeseriesanalysis #insights #conversationalai #selfhosted #userinsights #Chatbot #llms #regex #filters
chatbot analytics llm analysis llm monitoring llm observability Conversational Analytics user analytics llm domain conversations
มุมมอง: 18
วีดีโอ
PostHog-LLM Chat analytics: Analyzing Dialog Text Size Trends | HogQL
มุมมอง 373 หลายเดือนก่อน
In this video, I explored text size trends using PostHog-LLM. I visualize the average text length for both users and models utterances, compared phone vs desktop users, and see which ChatGPT version-3.5 or 4-generates longer responses. This video uses some HogQL expressions aswell :) Project: minuva.com/PostHog-LLM GitHub Page: github.com/postlang/posthog-llm Dataset: huggingface.co/datasets/al...
Tracking AI Toxicity: PostHog-LLM Analysis of WildChat 1M Dataset
มุมมอง 333 หลายเดือนก่อน
I analyze the toxicity trends of various OpenAI models using the WildChat 1M dataset, which spans from April 2023 to April 2024. By using PostHog-LLM and HogQL expressions, I compare the toxicity rates of both AI models and users over time. Project: minuva.com/PostHog-LLM GitHub Page: github.com/postlang/posthog-llm Dataset: huggingface.co/datasets/allenai/WildChat-1M Discord: discord.com/invit...
Using PostHog-LLM Filters to Analyze Conversations and Detect Patterns
มุมมอง 323 หลายเดือนก่อน
Demonstrating how to use PostHog-LLM to identify specific patterns in user dialogs with the filters feature. I explore both simple string searches and more complex use cases, such as email requests, email corrections, email replies, and more. Project: minuva.com/PostHog-LLM GitHub Page: github.com/postlang/posthog-llm Dataset: huggingface.co/datasets/allenai/WildChat-1M Discord: discord.com/inv...
PostHog-LLM: Using Dashboards to visualize conversational datasets
มุมมอง 1703 หลายเดือนก่อน
A quick overview of the Dashboards in PostHog-LLM. I share some insights from past videos (jailbreak prompts, user and agent toxicity, conversation timestamps, and funnel insights) and finally show how to create a dashboard and adda new insight. I’ll be answering questions like: What percentage of conversations begin with a jailbreaking prompt? What’s the overall trend over time? And I’ll also ...
PostHog-LLM: Analyzing Jailbreaking Dialogs with Meta's Prompt-Guard
มุมมอง 433 หลายเดือนก่อน
In this quick video, I show insights regarding jail breaking prompts using PostHog-LLM with the WildChat Conversational dataset. I’ll be answering questions like: What percentage of conversations begin with a jailbreaking prompt? What’s the overall trend over time? And I’ll also showcase some real examples of user jailbreak prompts. Project: minuva.com/PostHog-LLM GitHub Page: github.com/postla...
Chat Insights: Using Funnel Chart for Conversational Data | PostHog-LLM
มุมมอง 353 หลายเดือนก่อน
Discover how funnel charts can reveal insights into conversation turn times using the Wild Chat 1M conversational dataset. I explore how many users have just 1 turn, how many engage with more than 2 turns, and whether user engagement is increasing over time. Project: minuva.com/PostHog-LLM GitHub Page: github.com/postlang/posthog-llm Dataset: huggingface.co/datasets/allenai/WildChat-1M Discord:...
Chat Insights: Measuring Response Time Between Conversation Turns with PostHog-LLM | Funnel Analysis
มุมมอง 333 หลายเดือนก่อน
I demonstrate how to calculate the average time it takes for users to transition from the first to the second turn in a conversation. I also explore the distribution of these response times to get deeper insights from user behavior. Project: minuva.com/PostHog-LLM GitHub Page: github.com/postlang/posthog-llm Dataset: huggingface.co/datasets/allenai/WildChat-1M Discord: discord.com/invite/uNurWj...
Chat Insights: Identifying Users with Long Conversations using PostHog-LLM
มุมมอง 204 หลายเดือนก่อน
how to quickly identify users with long session times using PostHog's-LLM built-in features. This video provides a simple walk through of tracking session lengths and pinpointing users with extended interactions. Project: minuva.com/ PostHog-LLM GitHub Page: github.com/postlang/posthog-llm Dataset: huggingface.co/datasets/allenai/WildChat-1M Discord: discord.com/invite/uNurWjDkke #llm #analytic...
Chat Insights: Analyzing Conversation Duration with PostHog-LLM
มุมมอง 234 หลายเดือนก่อน
In this video, I'm diving into session duration stats from a conversational dataset (WildChat 1M). I show how to calculate the average session time for all conversations, display the monthly averages, and remove outliers (long conversations) to get more realistic view of your conversation duration. Project: minuva.com/ PostHog-LLM GitHub Page: github.com/postlang/posthog-llm Dataset: huggingfac...
PostHog-LLM: Analyzing Country Data Property | Conversational Analytics Breakdown
มุมมอง 194 หลายเดือนก่อน
I'll be diving into the WildChat-1M Conversational Dataset using PostHog-LLM! This time, I'll focus on analyzing the country property within the dataset. I'll answer questions like: Which are the most popular countries users come from? and And which countries are trending over time? I'll also see instances where users have changed countries mid-conversation, showing which countries they switch ...
Uncovering Anomalies in WildChat-1M Data Using PostHog-LLM | Conversational Data Analysis
มุมมอง 244 หลายเดือนก่อน
In this video, I dive back into exploring the WildChat-1M Conversational Dataset using PostHog-LLM. I demonstrate a simple example of an anomaly I discovered in the data and work through extracting more insights from it. Project: minuva.com/ PostHog-LLM GitHub Page: github.com/postlang/posthog-llm Dataset: huggingface.co/datasets/allenai/WildChat-1M Discord: discord.com/invite/uNurWjDkke #llm #...
Self Host PostHog-LLM Using CapRover in Your Infrastructure
มุมมอง 3404 หลายเดือนก่อน
In this video, I showcase how easily install locally PostHog-LLM (PostHog variant for text analytics extensions serving as an advanced LLM analytics) into CapRover. This guide walks you through the entire deployment process, from setting up your CapRover environment to setting up PostHog-LLM using CapRover's one-click apps template. Repository with instructions: github.com/AndrMoura/posthog-llm...
Complete PostHog Setup on CapRover Using DigitalOcean Droplets
มุมมอง 4824 หลายเดือนก่อน
Hi, in this video, I’ll guide you through the easy process of deploying PostHog on Digital Ocean using CapRover as a PaaS solution. This tutorial covers: - Creating an Ubuntu droplet - Installing Docker and CapRover - Configuring a Digital Ocean domain with an existing Namecheap domain - Setting up CapRover’s root domain and enabling HTTPS - Deploying PostHog using the one-click app template - ...
Deploy PostHog in Your Infrastructure with CapRover
มุมมอง 4064 หลายเดือนก่อน
Hello, today I showcase how easily install locally PostHog into CapRover. This guide walks you through the entire deployment process, from setting up your CapRover environment to setting up PostHog using CapRover's one-click apps template. Repository with instructions: github.com/AndrMoura/posthog-caprover Project: minuva.com/ PostHog-LLM GitHub Page: github.com/postlang/posthog-llm Discord: di...
PostHog-LLM: Chat data insights using timestamps & HogQL
มุมมอง 724 หลายเดือนก่อน
PostHog-LLM: Chat data insights using timestamps & HogQL
PostHog-LLM: Finding toxic users | Cohorts & Actions - Part 2
มุมมอง 225 หลายเดือนก่อน
PostHog-LLM: Finding toxic users | Cohorts & Actions - Part 2
PostHog-LLM: Finding toxic users | Cohorts - Part 1
มุมมอง 395 หลายเดือนก่อน
PostHog-LLM: Finding toxic users | Cohorts - Part 1
PostHog-LLM: Easy and Quick Insights Wildchat-1M Conversational Dataset
มุมมอง 135 หลายเดือนก่อน
PostHog-LLM: Easy and Quick Insights Wildchat-1M Conversational Dataset
PostHog-LLM: Domain breakdown your conversations using time series plot | WildChat-1M Dataset
มุมมอง 95 หลายเดือนก่อน
PostHog-LLM: Domain breakdown your conversations using time series plot | WildChat-1M Dataset
Using PostHog-LLM to Track User Dialog Trends And Categories | WildChat-1M Dataset
มุมมอง 116 หลายเดือนก่อน
Using PostHog-LLM to Track User Dialog Trends And Categories | WildChat-1M Dataset
PostHog-LLM: Extracting coding languages from conversational data
มุมมอง 226 หลายเดือนก่อน
PostHog-LLM: Extracting coding languages from conversational data
PostHog-LLM: How to setup A/B testing for your LLM application
มุมมอง 1588 หลายเดือนก่อน
PostHog-LLM: How to setup A/B testing for your LLM application
PostHog-LLM: Visualize your user dialogs within Trends
มุมมอง 268 หลายเดือนก่อน
PostHog-LLM: Visualize your user dialogs within Trends
PostHog-LLM: Measure User Engagement and Feature Adoption
มุมมอง 1458 หลายเดือนก่อน
PostHog-LLM: Measure User Engagement and Feature Adoption
PostHog-LLM: Measuring Failed User-Agent Interactions
มุมมอง 408 หลายเดือนก่อน
PostHog-LLM: Measuring Failed User-Agent Interactions
PostHog-LLM: Finding your power users | Stickiness
มุมมอง 1059 หลายเดือนก่อน
PostHog-LLM: Finding your power users | Stickiness
PostHog-LLM: Insights from toxic and non-toxic users | Cohorts
มุมมอง 1059 หลายเดือนก่อน
PostHog-LLM: Insights from toxic and non-toxic users | Cohorts
great explanation, thank you! I think I'll use Hetzner
awesome, thank you!
Hi Andra, Great stuff, thanks for the insight.. I want to install posthog locally on linux server so that it dont have any connection outside my intranet and all processing happen in our on prem. Please share the steps you followed
Hello, have you checked our repository, we have one click install command, easy to one click install. I also posted a video on how to install locally using CapRover: th-cam.com/video/acPLzzzcui8/w-d-xo.html
@@AndreMoura-lj6dd i will check and feedback thanks
Please upload on 1080p or 1440p.
Forgot to mention in the video: at 6:43 in the namecheap control panel, notice the nameservers. You need to add digital ocean's NS records!
Check out PostHog-LLM: github.com/postlang/posthog-llm Chat with ChatGPT and send it to PostHog-LLM: github.com/AndrMoura/posthog-llm-agent