Skip to main content

Posts

How to Nail Your Product Definition

Let's be honest, most product definitions suck. They're either packed with jargon that makes your eyes glaze over, filled with features nobody gives a crap about, or so vague they could be about anything. And most importantly, they totally miss the unfair advantage. Wait, what the hell is an unfair advantage?  Simply, it's the killer feature or a strategic edge that's so good, the others can't even copy it. It can be so simple as a dark mode, or an App Store feature to let competitors hook in. It's like building with Lego: you want that one foundational piece that's the base for everything else. Start with a simple square? Cool. But with the right unfair advantage, you can build it into a freaking skyscraper that everyone wants a piece of. Let me break down how I start to build new products. Step 1: Forget the "What," Focus on the "Why" (and How It Makes Users' Lives Easier) Simplified: Customer Problem > Fancy Features If you can

It's 2024, Hacking Your Way to Truly Useful Products

Last weekend I got again fed up by SaaS companies and their permanent "digital engagement" noise, so I canceled. You guess what really fed me up then? The extortion when I cancel my subscription, leading to mandatory, useless interrogation practice - surveys: "We want to understand why you want to cancel" with dozens of questions! Folks, when you DON'T understand why and when customers cancel and why your awesome product has more churn than all wholesale companies combined, then your complete product metrics (if any) are wrong and your product team needs to reevaluate how they build products. Yes, the tech startup bubble obsesses over user and customer engagement. Your investors tell you that, the "marketing" gurus, "influencers," and I don't know who else. They tell you to push notifications, implement annoying gamification,and endless "sticky" features, desperately trying to keep eyeballs locked on our products and services.

Rethinking Product Management: Flexibility and Customer Obsession for Success

I've been building products for a long time now, moving from Solution Engineer and Solution Architect over Product Manager to my current role as CPO. Along the way, I've seen the landscape shift dramatically. One thing's for sure: if you want to create products that customers truly love (and that drive real business results), you need to stay obsessed with their experience. That means rethinking some of the old "tried and true" ways of doing things. Don't: Just Adding Features Experience is critical for building customer loyalty: A great interface sells software. No great customer experience, no sales. Product management isn't about mindlessly churning out features. It starts with a deep understanding of your customers, the market and your competition. What drives the customer behavior? What are their biggest pain points?  To answer those questions, you need a toolkit that includes research, analytics, and direct feedback channels. This empathy for your cu

AI's False Reality: Understanding Hallucination

Artificial Intelligence (AI) has leapfrogged to the poster child of technological innovation, on track to transform industries in a scale similar to the Industrial Revolution of the 1800s. But in this case, as cutting-edge technology, AI presents its own unique challenge, exploiting our human behavior of "love to trust", we as humans face a challenge: AI hallucinations. This phenomenon, where AI models generate outputs that are factually incorrect, misleading, or entirely fabricated, raises complex questions about the reliability and trust of AI models and larger systems. The tendency for AI to hallucinate comes from several interrelated factors. Overfitting – a condition where models become overly specialized to their training data – can lead to confident but wildly inaccurate responses when presented with novel scenarios (Guo et al., 2017). Moreover, biases embedded within datasets shape the models' understanding of the world; if these datasets are flawed or unrepresen

When to Choose ETL vs. ELT for Maximum Efficiency

ETL has been the traditional approach, where data is extracted, transformed, and then loaded into the target database. ELT flips this process - extracting data and loading it directly into the system, before transforming it. While ETL has been the go-to for many years, ELT is emerging as the preferred choice for modern data pipelines. This is largely due to ELT's speed, scalability, and suitability for large, diverse datasets generated by multiple different tools and systems, think about CRM, ERP datasets, log files, edge computing or IoT. List goes on, of course.. Data Engineering Landscape Data engineering is the new kind of DevOps. With the exponential growth in data volume and sources, the need for efficient and scalable data pipelines and therefore data engineers has become the new standard . In the past, limitations in compute power, storage capacity, and network bandwidth made the famous 3-word "let's move data round" phrase Extract, Transform, Load (ETL) the

Life hacks for your startup with OpenAI and Bard prompts

OpenAI and Bard   are the most used GenAI tools today; the first one has a massive Microsoft investment, and the other one is an experiment from Google. But did you know that you can also use them to optimize and hack your startup? Even creating pitch scripts, sales emails, and elevator pitches with one (or both) of them helps you not only save time but also validate your marketing and wording. Curios? Here a few prompt hacks for startups to create / improve / validate buyer personas, your startups mission / vision statements, and USP definitions. Introduce yourself and your startup Introduce yourself, your startup, your website, your idea, your position, and in a few words what you are doing to the chatbot: Prompt : I'm NAME and our startup NAME, with website URL, is doing WHATEVER. With PRODUCT NAME, we aim to change or disrupt INDUSTRY. Bard is able to pull information from your website. I'm not sure if ChatGPT can do that, though. But nevertheless, now you have laid a grea

Indexing PostgreSQL with Apache Solr

Searching and filtering large IP address datasets within PostgreSQL can be challenging. Why? Databases excel at data storage and structured queries, but often struggle with full-text search and complex analysis. Apache Solr, a high-performance search engine built on top of Lucene, is designed to handle these tasks with remarkable speed and flexibility. What do we need? A running PostgreSQL database with a table containing IP address information (named "ip_loc" in our example). A basic installation of Apache Solr. Setting up Apache Solr Create a Solr Core: Bash solr create -c ip_data -d /path/to/solr/configsets/ Define the Schema ( schema.xml ) XML < field name = "start_ip" type = "ip" indexed = "true" stored = "true" /> < field name = "end_ip" type = "ip" indexed = "true" stored = "true" /> < field name = "iso2" type = "string" indexed = "true&q

Some fun with Apache Wayang and Spark / Tensorflow

Apache Wayang is an open-source Federated Learning (FL) framework developed by the Apache Software Foundation. It provides a platform for distributed machine learning, with a focus on ease of use and flexibility. It supports multiple FL scenarios and provides a variety of tools and components for building FL systems. It also includes support for various communication protocols and data formats, as well as integration with other Apache projects such as Apache Kafka and Apache Pulsar for data streaming. The project aims to make it easier to develop and deploy machine learning models in decentralized environments. It's important to note that this are just examples and they may not be the way for your project to interact with Apache Wayang, you may need to check the documentation of the Apache Wayang project ( https://wayang.apache.org ) to see how to interact with it. I just point out how easy it is to use different languages to interact between Wayang and Spark. Also, you need to mak

Get Apache Wayang ready to test within 5 minutes

Hey followers, I often get ask how to get Apache Wayang ( https://wayang.apache.org ) up and running without having a full big data processing system behind. We heard you, we built a full fledged docker container, called BDE (Blossom Development Environment), which is basically Wayang. Here's the repo:  https://github.com/databloom-ai/BDE I made a short screencast how to get it running with Docker on OSX, and we also have made two hands-on videos to explain the first steps. Let's start with the basics - Docker. Get the whole platform with: docker pull ghcr.io/databloom-ai/bde:main At the end the Jupyter notebook address is shown, control-click on it (OS X); the browser should open and login you automatically: Voila - done. You have now a full working Wayang environment, we prepared three notebooks to make it more easy to dive into. Watch our development tutorial video (part 1) to get a better understanding what Wayang can do, and what not. Click the video below: 

Combined Federated Data Services with Blossom and Flower

When it comes to Federated Learning frameworks we typically find two leading open source projects - Apache Wayang [2] (maintained by  databloom ) and Flower [3] (maintained by  Adap ). And at the first view both frameworks seem to do the same. But, as usual, the 2nd view tells another story. How does Flower differ from Wayang? Flower is a federated learning system, written in Python and supports a large number of training and AI frameworks. The beauty of Flower is the strategy concept [4]; the data scientist can define which and how a dedicated framework is used. Flower delivers the model to the desired framework and watches the execution, gets the calculations back and starts the next cycle. That makes Federated Learning in Python easy, but also limits the use at the same time to platforms supported by Python.  Flower has, as far as I could see, no data query optimizer; an optimizer understands the code and splits the model into smaller pieces to use multiple frameworks at the same ti

Compile Apache Wayang on Mac M1

We release Apache Wayang  v0.6.0 in the next days, and during the release testing I was wondering if we get wayang on M1 (ARM) running. And yes, a few small changes - voila! Install maven, scala, sqlite and groovy: brew install maven scala groovy sqlite Download openJDK 8 for M1: https://www.azul.com/downloads/?version=java-8-lts&os=macos&architecture=arm-64-bit&package=jdk  and install the pkg.  Get Apache Wayang either from  https://dist.apache.org/repos/dist/dev/wayang/ , or git-clone directly: git clone https://github.com/apache/incubator-wayang.git Start the build process: cd incubator-wayang export JAVA_HOME=/Library/Java/JavaVirtualMachines/zulu-8.jdk/Contents/Home mvn clean install Ready to go: [INFO] Reactor Summary for Apache Wayang 0.6.0-SNAPSHOT: ... [INFO] BUILD SUCCESS [INFO] ------------------------------------------------------------------------ [INFO] Total time:  06:24 min After the build is done the binaries are located in mavens home: ~/.m2/repository/o

Stream IoT data to S3 - the simple way

First, a short introduction to infinimesh , an Internet of Things (IoT) platform which runs completely in Kubernetes :  infinimesh enables the seamless integration of the entire IoT ecosystem independently from any cloud technology or provider. infinimesh easily manages millions of devices in a compliant, secure, scalable and cost-efficient way without vendor lock-ins. We released some plugins over the last weeks - a task we had on our roadmap for a while. Here is what we have so far: Elastic Connect infinimesh IoT seamless into Elastic . Timeseries Redis-timeseries with Grafana for Time Series Analysis and rapid prototyping, can be used in production when configured as a Redis cluster and ready to be hosted via Redis-Cloud . SAP Hana All code to connect infinimesh IoT Platform to any SAP Hana instance Snowflake All code to connect infinimesh IoT Platform to any Snowflake instance. Cloud Connect All code to connect infinimesh IoT Platform to Public Cloud Provider AWS, GCP and Azu

Embedded Linux won't reboot - how to fix and repair

I have a lot of embedded systems running in our lab or in my home, all of them either as Raspberries or selfmade PCB with Yocto. Sometimes I can't reboot some systems, I think its the journald which causes some issues with SSD Cards, the error-message usually is: Failed to open /dev/initctl Anyhow, if you have this issue - a reboot can be force-forced: systemctl --force --force reboot Since a forced reboot does not sync the journal, the system typically comes up with a damaged FS. The remote fsck can be initiated by extending the command above with sudo tune2fs -i 1m /dev/DISK && touch /forcefsck && systemctl --force --force reboot (assumed you have access to a shell, via SSH or local access). When all goes fine, the system comes up with a clean FS. All this fuss comes from the SSD r/w actions, a well designed IoT embedded devices should have a flash mem part for the running OS.