Jack Walker

We bring qualified leads for your business through Facebook ads and organic social media marketing | Helped 50+ clients with profitable paid campaigns (5x ROI) | Award-winning digital marketing agency🚀

Do your staff spend hours trying to create social media content to generate sales/ leads? 

Do you want to make more money from social media? 

Are the posts not getting engagements? 

Do you want your posts to go out to targeted profiles!?  

Yes it’s possible!!! 

We have your solution for you! OptiModo Marketing has the answers to your problems!

We create custom social media marketing projects for YOU, focused on your ROI (Return On Investment) making sure you get the most out of your marketing!

Did you know the average company spends an hour a day posting on their social media channels? That means you have spent roughly (if you pay them £15 an hour) £450 a month on social media. 

Our social media management packages start at £219 saving you time and money. 

We also make sure your content is high quality and being seen by your target market. 

The best thing about us is THERE ARE NO LOCK IN CONTRACTS! We will send you a report after each month, and if you are not happy (we are sure you will be) you can leave!

Not only do we do social media management, we can make sure your website is up to date, make sure it is working to its potential. We can design, build or modify your website to your needs, with your budget.

-------------------------------------

What can we help with?
Social Media Marketing
Social Media Management
Social Media Content
B2B Leads
B2C Leads
Website Design
Website Maintenance 
And much more!

Managing Director

Enabling businesses to unlock their full potential using social media, websites and email marketing.

Stopping you SPENDING money on your marketing and helping you MAKE money marketing.

Data & AI Enthusiast | Attracting High-Quality Opportunities Through Personal Branding & Smart Outreach | German Expat Mastering Dutch Life

I’m Francie (you pronounce it like “fancy”) :P, the Head of Global Sales Development who wants to turn the bad image of SDRs around through personal branding, content creation, and smart outbound. 

If you’re curious, here’s a quick rundown:

👋 Who am I? 
A bubbly, face-to-face extrovert. My background? Everything from big corporations to tiny AI & Data startups. Now, I’m on a mission to translate that energy into driving growth online. 

🤝 What do I do? 
I help businesses create more high-quality opportunities using savvy outbound strategies, lead-generating content, and genuine personal branding. Fed up with outdated tactics? You’re in the right place!

Francie Kastl

Kafka had its moment. But real-time analytics has evolved -- and Kafka hasn't kept pace.

The problem isn't that Kafka is bad. It's that Kafka was built for event-driven architectures, not modern real-time analytics. Record-based storage means you're moving entire records across the network even when you only need a few columns. No native schemas. No columnar optimization. No unified lakehouse integration.

Apache Fluss changes the game.

Built for the lakehouse era, Fluss delivers what real-time analytics actually needs:
→ Columnar storage with projection pushdown (5x compression vs Kafka)
→ Schematized tables with PK support for mutable data
→ Zero-copy tiering between streaming, warm, and cold layers
→ Native Paimon/Iceberg integration

The result? Lower storage costs, reduced network traffic, and unified access to real-time and historical data without complex duplication pipelines.

The third wave of streaming data is here. Read our latest on why the future of real-time analytics looks different from the past decade.

Find the link to the respective blog article of my colleague Alex Campos in the comments.

If your “real-time AI” runs on batch data from yesterday, we need to talk. | Head of Global Sales Development | German potato enjoying the vibrant Netherlands :)

𝐀𝐩𝐚𝐜𝐡𝐞 𝐅𝐥𝐮𝐬𝐬: 𝐓𝐡𝐞 𝐆𝐚𝐦𝐞 𝐂𝐡𝐚𝐧𝐠𝐞𝐫 𝐢𝐧 𝐃𝐚𝐭𝐚 𝐒𝐭𝐫𝐞𝐚𝐦𝐢𝐧𝐠

Over the past few months, I’ve been diving into Apache Fluss (Incubating) and building a Streamhouse powered by Fluss, Flink, and Iceberg. This architecture processes bank transactions in real time, detects fraud, and serves data seamlessly across hot (sub-second latency) and cold (minutes latency) layers.

𝐖𝐡𝐚𝐭 𝐢𝐬 𝐀𝐩𝐚𝐜𝐡𝐞 𝐅𝐥𝐮𝐬𝐬?

Apache Fluss is an open-source streaming storage system designed to cut pipeline costs and bridge the gap between real-time streaming and the data lakehouse. It enables real-time applications, analytics and OLAP workloads.

𝐇𝐨𝐰 𝐈𝐭 𝐖𝐨𝐫𝐤𝐬

Fluss runs on a distributed, fault-tolerant cluster with two main components:

• CoordinatorServer: Handles metadata, permissions, scaling, failover, and table management.
• TabletServer: Stores and serves data via:
 - LogStore: Append-only logs (also acts as a changelog and recovery mechanism), backed by Apache Arrow for column pruning during streaming reads.
 - KvStore: Tables with updates and deletes, backed by RocksDB for high-performance lookups.
Data is organized into buckets and tablets:
• LogTablet: Stores log data for both Log and Primary Key tables.
• KvTablet: Stores primary table data for updateable tables.
Remote storage offloads LogStore data and snapshots KvStore for fault tolerance and disk cost saving. ZooKeeper currently manages coordination.

𝐓𝐚𝐛𝐥𝐞 𝐓𝐲𝐩𝐞𝐬

Fluss supports two table types:

• Log Tables: Append-only, ideal for high-throughput logs (similar to Kafka). They preserve write order and support column pruning for faster queries.
• Primary Key Tables: Support INSERT, UPDATE, DELETE, enforce uniqueness, and allow partial column updates. Changes are captured in a changelog for streaming reads.

𝐋𝐚𝐤𝐞𝐡𝐨𝐮𝐬𝐞 𝐈𝐧𝐭𝐞𝐠𝐫𝐚𝐭𝐢𝐨𝐧

Fluss is not just streaming storage, it’s the foundation for a Streaming Lakehouse, also named Streamhouse.

Fluss Cluster stores recent data in Arrow format for Log Tables, optimized for low-latency reads/writes (days of data, sub-second freshness).

Lakehouse Storage compacts data using open table formats (Paimon, Iceberg, Lance), optimized for analytics and long-term retention, months of data with minute-level freshness, through the Flink-based tiering service.

Fluss offers unified metadata for both streaming and historical data, no duplication or inconsistency.

Via Union Reads (currently only) Flink can query real-time + historical data together, enabling true real-time analytics.

Fluss synchronizes metadata with lake catalogs, enabling engines like Spark, Flink, Trino, and Dremio to directly read open table formats.

The legacy Streaming Data Era is coming to an end 👋 As real-time analytics demands outgrow Kafka’s log-based model, it’s clear that we need a storage layer built for analytics, not just messaging🔝

The future of stream processing is about actionable data now, and Apache Fluss is paving the way for next-gen Streaming Applications, making fresh data queryable and real-time insights actually real.

Learn what a world without Kafka looks like. Link in the first comment.

Kafka Didn’t Fail. Real-Time Analytics Moved On.

Kafka was never meant to be an analytics platform - yet that’s exactly what many teams are forcing it to become.

For over a decade, Kafka powered event-driven architectures by doing one thing extremely well: moving events reliably at scale. That achievement deserves respect. But today’s real-time analytics requirements expose a hard truth:

Kafka is being stretched far beyond its original purpose.

The Core Problem: Transport vs. Insight
Kafka is a record-based, append-only log. That design is ideal for ingestion pipelines and microservice communication - but it’s fundamentally misaligned with analytics workloads.

Analytics engines don’t want full records.
They want columns, filters, aggregations, and state.

Because Kafka can’t provide this natively, real-time analytics architectures built on Kafka inevitably accumulate:
-External processing engines pulling full records
-Schema registries to compensate for missing governance
-Duplicate data across hot, warm, and cold layers
-Rising network, storage, and operational costs

If your “real-time” platform requires three systems just to answer a query, something is wrong.

Streaming Is in Its Third Wave
>The first wave of streaming was about movement. Kafka won.
>The second wave added processing - enrichment, aggregation, complex events.
>The third wave is about decisioning: real-time analytics, AI pipelines, and Lakehouse-native architectures. In this world:
Streams behave like tables
State is first-class
Column-level efficiency matters
Real-time and historical data must be unified

Kafka’s record-centric model becomes the bottleneck.

Why Fluss Changes the Equation
Apache Fluss isn’t “Kafka, but faster.” It’s a different assumption entirely.
Fluss treats streams as schematized tables, stored in columnar format (Apache Arrow), enabling:
>Column pruning and predicate pushdown
>Dramatically lower network and storage overhead
>Native support for mutable, stateful data via primary key tables
>No external key-value stores. No constant full-record scans.

Built for the Lakehouse Era
Kafka was born in the Data Lake era. Fluss was designed for the Lakehouse.
With built-in tiering and zero-copy data aging into object storage and Lakehouse formats, Fluss eliminates the need to duplicate streams just to make them analyzable.

Paired with Apache Flink, it forms a foundation for true real-time analytics, not just real-time transport.

This Isn’t Anti-Kafka. It’s Pro-Evolution.
Kafka still has an important role. But using it as the backbone of real-time analytics is like using a message queue as a database - it works until scale, cost, and complexity catch up.

The next generation of streaming architectures won’t be defined by how fast data moves > but by how quickly it becomes insight.

And that requires a new kind of storage.

From Lakehouse to Streamhouse: Real-Time Without the Mess

Yesterday we hosted a webinar on “Evolving to Streamhouse: Apache Fluss × Iceberg” with contributors from across the lakehouse and streaming ecosystems. 

One question kept coming up:

👉 How do we get sub-second analytics without duplicating data, pipelines, or operational pain?

The short answer: the lakehouse needs to become streaming-native.

🧊 Iceberg is powerful, but not real-time by design.
 It excels at open, reliable analytics on object storage. But when freshness matters (fraud, ops, personalization, AI), minute-level latency pushes teams toward extra systems… and suddenly you’re back to a modern Lambda architecture.

🔥 Enter Streamhouse.

 A unified architecture with:

>One write path
>One read path
>One logical table

With Apache Fluss as the real-time layer and Iceberg as the historical layer, hot data stays fast, cold data stays cheap and queries always see the latest truth via union reads.

⚡ Zero-ETL tiering means no fragile ingestion pipelines.
 🧠 Deletion vectors across hot + cold layers make real-time upserts correct at scale.
 🔄 Flink CDC becomes the front door, streaming entire databases with schema evolution handled automatically.

The result?

 📊 Real-time dashboards
 🤖 Fresher features for ML & AI
 📈 Historical analytics
 -all on the same tables, with consistent semantics.

Streamhouse doesn’t replace Iceberg.
It extends it - bringing the lakehouse into the real-time era.

🎥 Missed the session? Watch the full recording here 👇

https://lnkd.in/dCAygn84

Fresha started where most companies start: CDC into Snowflake, dbt models, the familiar data stack. 

As the platform evolved to meet real-time requirements, operational analytics, and sub-second latency demands, it required a massive redesign. What emerged was something that looked remarkably like a streaming lakehouse, also known as Streamhouse, built piece by piece out of necessity. 

Join us for a behind-the-scenes look at Fresha's evolution of its data platform.

Giannis Polyzos, Principal Streaming Architect at Ververica, and Anton Borisov, Principal Data Architect at Fresha, will cover:

- Why Fresha moved from batch to streaming lakehouse
- The architectural tradeoffs that hurt (and the ones that paid off)
- The tools we chose and the reasoning behind them
- How Apache Fluss fills the critical gaps between streaming and lakehouse systems
- Practical lessons for teams on the same journey

Fresha is the UK's first StarRocks user, and now Apache Fluss is next for them.

Attend here: https://lnkd.in/ePadrQ-K

Ververica | Original creators of Apache Flink®

View company: Ververica | Original creators of Apache Flink®

Fresha started where most companies start: CDC into Snowflake, dbt models, the familiar data stack. As the platform evolved with real-time requirements, operational analytics, and sub-second latency demands, it required a massive redesign, and what emerged was something that looked remarkably like a streaming lakehouse, also known as Streamhouse, built piece by piece out of necessity. Join us for a behind-the-scenes look at Fresha's data platform evolution.

Giannis Polyzos, Principal Streaming Architect at Ververica, and Anton Borisov, Principal Data Architect at Fresha, will cover:

- Why Fresha moved from batch to streaming lakehouse
- The architectural tradeoffs that hurt (and the ones that paid off)
- The tools we chose and the reasoning behind them
- How Apache Fluss fills the critical gaps between streaming and lakehouse systems
- Practical lessons for teams on the same journey

Fresha is the UK's first StarRocks user, and now Apache Fluss is next for them.

Maciej Mojsiewicz, Jan Siekierski and 7 other attendees

View event: Webinar: Whole Lotta Stream. The Streamhouse Architecture at Fresha

Webinar: Whole Lotta Stream. The Streamhouse Architecture at Fresha

taplio

The Battle

Jack Walker

Francie Kastl

Followers

Posts

Engagements

Likes

Performance ScoreFrancie Kastl's LinkedIn Performance

Ready to grow your LinkedIn presence?

taplio

The Battle

Jack Walker

Francie Kastl

Followers

Posts

Engagements

Likes

Performance Score.css-goub6s{font-size:var(--chakra-fontSizes-md);font-weight:500;color:var(--chakra-colors-gray-600);display:block;margin-top:var(--chakra-space-1);}Francie Kastl's LinkedIn Performance

Ready to grow your LinkedIn presence?

Performance ScoreFrancie Kastl's LinkedIn Performance