Scalytics Enablement

Kafka training for streaming and agent teams.

Q: Who teaches the cohorts?

Mirko Kämpf leads the Confluent programs. He is a data platform architect and Confluent specialist who designs streaming systems for enterprise teams. Scalytics-native tracks for KafScale, KafClaw, KafGraph, and KafSIEM are taught by the engineers who built the systems.

Kafka training that goes deeper than the bootcamp. Three weeks, three roles, hands-on.

Most streaming and agent platforms ship with documentation and a three-day bootcamp. We ship with a cohort program that runs alongside your team for the time it actually takes to learn the system. Architects, developers, and operators train together so they leave speaking the same language.

See open cohorts Talk to the enablement team

Format

Live cohort, 15 sessions of 90 minutes, hands-on labs, recordings included.

Pacing

Three to eight weeks, your team's pace, not ours.

Delivery

EU hosted, GDPR compliant, end-to-end encrypted learning workspace.

Three weeks · Three roles · Hands-on

Kafka training that goes deeper than the default bootcamp.

Role-paired cohorts for Confluent Kafka Platform and Cloud teams. Architects, developers, and operators train together. Delivered as a Confluent partner.

See open cohorts →

How cohorts stick

Learning by capturing ideas, not by watching slides.

Replays have a 5 percent completion rate. Boot camps fade in three weeks. The pattern that lasts is the one researchers have known about for decades: structured doing, articulation in the learner's own words, deliberate reflection, and connection to prior knowledge. Every Scalytics cohort runs on it.

Step 01

Do

Work through structured tasks on a role-specific path. Real labs against real systems, not toy clusters.

Step 02

Capture

Articulate each idea in your own words, spoken or written. This is the step that turns a tutorial into knowledge.

Step 03

Reflect

A coaching companion deepens understanding and sharpens questions. Human coaches verify and extend in every live session.

Step 04

Connect

Every concept links to a growing knowledge graph with public references. Follow any thread to Wikipedia or curated video.

Methodology delivered on maindset.ACADEMY, Scalytics' sovereign learning platform.

Kafka training · Open cohorts

Programs running now.

CFLT-CPOB-V1

Confluent Platform Onboarding

Role-based onboarding for teams running Confluent Platform on premises or on Kubernetes. Three weeks, three tracks, one shared foundation.

Duration

3 weeks

Sessions

15 live

Format

90 min each

Week 1. Foundation for all roles. Kafka architecture, Schema Registry, Shift Left, Flink preview.
Week 2. Developer and Architect tracks in parallel. Producer and consumer tuning, Kafka Streams DSL, scalability, HA, Event Sourcing, CQRS.
Week 3. Operator track. CFK on Kubernetes, monitoring, scaling, diagnostics, disaster recovery.

View program & enroll Delivered on maindset.ACADEMY

CFLT-CCOB-V1

Confluent Cloud Training

Same structure, optimized for teams running Confluent Cloud. Cloud-specific labs cover RBAC, networking, governance, and managed Flink.

Duration

3 to 8 weeks

Sessions

15 live

Format

90 min each

Week 1. Foundation for all roles. Cloud-native Kafka, Schema Registry, Shift Left, managed Flink preview.
Week 2. Developer and Architect tracks. Cloud producer and consumer tuning, Kafka Streams, Flink SQL deployment, cloud scaling and security patterns.
Week 3. Operator track. Cluster lifecycle on Confluent Cloud, observability, cost controls, diagnostics, disaster recovery.

View program & enroll Delivered on maindset.ACADEMY

Three roles, one program

Architect, Developer, Operator. Trained together.

Other vendors ship role-specific tracks that produce specialists who can't talk to each other. We ship one program with three roles running in parallel, so your architect, developers, and operators leave with shared vocabulary and shared muscle memory.

Role 01

Architect

For the person designing the streaming and agent layer. Week 1 foundation, Week 2 afternoons, three sessions deep dive.

Cluster sizing, scalability, and high availability patterns
Data product design and stream contracts
Shift Left and security architecture
Event Sourcing, CQRS, and Medallion patterns
Federation strategies for multi-cluster topologies

Role 02

Developer

For engineers shipping producers, consumers, and stream processors. Week 1 foundation, Week 2 mornings, three sessions deep dive.

Producer and consumer tuning, delivery semantics
Kafka Streams DSL and stateful processing
Flink SQL deployment and query patterns
Schema evolution and Schema Registry workflows
Error handling, retries, and DLQ patterns

Role 03

Operator

For the team running clusters in production. Week 1 foundation alongside others, Week 3 dedicated, three sessions deep dive.

CFK and cluster lifecycle on Kubernetes
Monitoring, observability, and capacity planning
Scaling brokers and rebalancing partitions
Diagnostics, incident response, and postmortems
Disaster recovery and multi-region failover

Pacing

Your team's calendar, not ours.

Typical vendor bootcamp

Three days, full-time, fixed.

Time off the desk. Three consecutive days where production stops being supported.
Front-loaded. Information density that no team retains past week two.
No follow-through. The voucher expires; the muscle memory never forms.
Role-siloed. Developers, architects, and operators in separate rooms or separate cohorts.
No feedback path. Issues surfaced in the workshop go nowhere useful.

Scalytics Enablement

Three to eight weeks, your pace.

90-minute live sessions. Fits around standups and on-call rotations.
Two pacing options. Four weeks intensive or eight weeks extended; pick what matches your team's bandwidth.
Lab time between sessions. Capture and reflect built into the schedule, not bolted on after.
Role-paired, not role-siloed. Foundation week is shared so terminology converges across the team.
Workshop to Engineering feedback path. Issues surfaced in cohorts feed Scalytics Engineering directly. A bug a participant hits on Tuesday can land in a sprint on Friday.

Why Scalytics

Authority where it matters, scale where it doesn't.

Built by the people who built Wayang.

Scalytics was founded by the original creators of Apache Wayang, an Apache Top-Level Project for federated data processing. The same authority that puts Databricks behind Spark and Confluent behind Kafka. The instructors teaching streaming patterns are also the ones writing the federation engine those patterns run on.

Engineering feedback loop, not a help desk.

Confluent and Cloudera are too big for a participant's friction to reach a product engineer. We're not. A diagnostic question raised in a Tuesday workshop can become a Friday sprint item. Enterprises buying enablement get more than training; they get a back-channel into the platform team.

Complementary to Confluent, not a substitute for it.

Scalytics is a Confluent partner. Our cohorts make teams that bought Confluent Platform or Confluent Cloud more fluent than the vendor's own three-day model produces. Confluent certifications still apply; our programs prepare candidates well past the exam-style baseline.

Roadmap

Tracks shipping through 2026.

The Kafka training programs are first because that's where most enterprises start, anchored on the Confluent ecosystem our team has worked in for years. Scalytics-native tracks come next, each tied to a product our team built and ships in production.

Q3 2026 · Operator

KafScale Operator

Run Kafka workloads stateless on S3, no rebalance storms, no broker babysitting. For operators leaving storage-coupled clusters behind.

Q3 2026 · Developer

KafClaw Agent Developer

Build agent runtimes on a policy mesh, not on brittle frameworks. For developers shipping the full LLM, tool, and state loop.

Q4 2026 · Architect

KafGraph Architect

Binary knowledge graph on Kafka with auto-expiring facts. For architects building agent memory vector databases cannot deliver.

Q4 2026 · Analyst

KafSIEM Analyst

Entity graphs for air-gapped operations with immutable provenance on every edge. For analysts in defense and critical infrastructure.

Want first access when cohorts open? Get on the enablement list.

Private cohorts

Bring this to your team.

Most engagements are private cohorts customized to the customer's stack and pacing. The curriculum stays, the labs and case material adapt to the platform you actually run. Typical engagement is one team of 6 to 18 engineers across the three roles.

Talk to the enablement team See the learning platform

Common questions

Before you enroll.

Who is this for?

Engineering teams running Confluent Platform, Confluent Cloud, Apache Kafka, or planning to. Architects, developers, and operators are all addressed in one cohort, so the whole team can attend together. Solo learners are welcome but the program is structured for groups.

How is this different from a vendor bootcamp?

Most Kafka training is a three-day, full-time, role-specific bootcamp. Our cohorts are three to eight weeks, 90-minute sessions, with three roles running in parallel and a shared foundation week. Scalytics is a Confluent partner, so our programs complement Confluent's certifications and prepare candidates beyond the exam baseline.

Does this prepare for Confluent certifications?

Yes. The curriculum covers the topic areas in CCDAK (Developer), CCAAK (Administrator), and CCAC (Cloud Operator). The certifications themselves are taken on Confluent's platform; we don't proctor exams.

What languages are sessions delivered in?

English and German. Private cohorts can be scheduled in either language. Course materials are bilingual where it improves clarity for cross-border teams.

What are the prerequisites?

Working familiarity with at least one general-purpose programming language. The Operator track adds basic Kubernetes familiarity. The Architect track assumes existing experience designing distributed systems. The Developer track is the most accessible to engineers new to streaming.

Where is learner data stored?

In the EU. The learning platform is GDPR compliant and end-to-end encrypted. No learner content leaves the EU at any point during the program.

Do we get recordings and materials after the cohort ends?

Yes. Session recordings, the concept network, shared docs, and lab materials stay accessible to the cohort after the program closes. Captured notes belong to the learner.

How is pricing structured?

Per-seat for public cohorts. Flat-fee for private cohorts, scaled by team size and customization. We don't sell vouchers that expire unused; pricing assumes the team actually completes the program. Talk to the enablement team for a quote.

Who teaches the cohorts?

Mirko Kämpf leads the Confluent programs. He's a data platform architect and Confluent specialist who designs streaming systems for enterprise teams. Scalytics-native tracks (KafScale, KafClaw, KafGraph, KafSIEM) are taught by the engineers who built the systems.

The experts for mission-critical infrastructure.

Launch your data + AI transformation.

Book your free discovery call now!