From Sensor to Stream Processing: Leadership and Culture Insights from “Manuel Frech, Teamlead Hard- & Firmware bei smaXtec” (smaXtec animal care GmbH)

Executive overview: A full-stack product organization with real-world impact

In “Manuel Frech, Teamlead Hard- & Firmware bei smaXtec” from smaXtec animal care GmbH, we get a precise look at an R&D organization that connects hardware, firmware, cloud infrastructure, backend, data science, and frontend into one product flow. The distinguishing feature: the value chain literally starts inside the cow and ends with metrics displayed in a browser tool and mobile apps. That end-to-end responsibility shapes team structure, project execution, hiring, onboarding, and quality.

This recap outlines what engineers can learn from the session—and why this organization is compelling for talent that wants to work on tangible products powered by a modern data pipeline.

Five teams, one product: Clear responsibilities, seamless handoffs

Manuel Frech breaks down the development department into five groups that jointly deliver the product:

• Hardware team: “We capture the data in the cow and then send the data out.” This team also owns firmware—“the program that runs on the microcontroller.”
• Infrastructure team: Ensures that “the servers and the entire network infrastructure” are in place “to receive all the data.”
• Backend team: Stores the data and provides interfaces “for internal and external” use.
• Data science team: Processes the data “through various algorithms to develop interesting metrics from raw data,” including signal processing algorithms, machine learning, and neural networks.
• Frontend team: “Displays the evaluated data”—both in the browser tool and the iOS/Android apps.

The key takeaway: every discipline contributes to a single product, and each group owns its part of the end-to-end flow. Ownership and clarity run across the stack, from physical sensing to visual output.

Matrix organization: Cross-team projects are standard

“Our projects are mostly cross-team,” Frech explains. To manage this, “a matrix organization” spans the entire dev team. Practically, that means:

• Every project includes contributors from each technical team.
• Cross-functional collaboration is the norm, not the exception.
• The matrix structure keeps complex, multi-discipline roadmaps manageable.

For engineers, this implies there’s no siloed work. You feel the impact of your contributions at the handoffs—whether bridging firmware to cloud ingestion, backend to data pipelines, or all the way through to frontend.

Hiring: Fast, collaborative, and intentionally relaxed

Once the need for a new team member is clear, the team lead and HR operate as a tight unit. The flow looks like this:

1. Define the role profile with HR.
2. Publish the job post.
3. Jointly review applications and invite promising candidates.
4. Conduct interviews with “the HR first point of contact” and “the Technical Team Lead.”

The tone is purposeful: “We always try to create a very relaxed atmosphere, which we generally cultivate in the company.” The goal is authenticity: “You need a relaxed atmosphere to start the conversation and show yourself as you really are. That is our main goal.” During the conversation, the team “teases out social and technical qualifications” to form “an overall picture of the person.”

For candidates, this underscores what matters most: be yourself, articulate your technical thinking, and engage as a collaborator. Social competence is part of the evaluation, not an afterthought.

Onboarding: Buddy from hour one, 2–3 weeks of context, and domain immersion

Once an offer is accepted, onboarding emphasizes guidance and context:

• Buddy system from the first hour: “This buddy is the responsible person for all questions. You can go to them, no matter what it is about.”
• A two- to three-week onboarding phase: You get an overview of all departments.
• Domain entry: A “small training” provides a direct introduction “into dairy farming” for background.
• Close technical support: “Tight technical mentoring in the first weeks is very important so that you can find your way into the area you’re working in.”

This is more than checklists and system access. New joiners learn how the whole product and domain fit together—from sensor logic to data processing to practical use in dairy farming.

Hardware and firmware: Beyond electronics—product, production, quality

Frech underscores that the hardware team’s remit is broad:

• Firmware engineering: “The program that runs on the microcontroller.” Ideal for low-level software enthusiasts.
• Full product responsibility: “We don’t just develop the electronics itself; it’s about the entire product—the finished sensor as it stands.”
• Production oversight: “We also monitor production. The entire production process also comes from us.”
• Quality assurance: “Quality monitoring is provided by us.”

For engineers who value end-to-end ownership, this is compelling. Work doesn’t stop at the schematic—it reaches into production processes and quality monitoring, forming the foundation for a reliable data chain.

From barn to cloud: Wireless ingestion, MQTT, Kafka, stream processing, ML/NN

The technical chain Frech outlines is modern and coherent, showing how hardware, infrastructure, backend, data science, and frontend interlock:

• Data capture: “We capture the data in the cow” and transmit wirelessly to reading devices.
• Cloud ingestion: From the reading devices, data goes “via MQTT into the cloud.”
• Stream processing: “There they are passed via Kafka into stream processing.”
• Data science enrichment: The team processes data using “signal processing algorithms, machine learning, and neural networks.”
• Visualization and use: Values are displayed in the frontend—“in our browser tool” and “in the iOS and Android app.”

The historical arc matters too: initially, the product ran with an on-premise solution—data “at the customer locally.” The company “decided fairly early to go to the cloud,” and for “about ten years” the data has been there. That shift signals a commitment to scalability and connected processing.

Culture that enables collaboration: Relaxed tone, clear ownership, real delivery

The alignment between structure and culture stands out:

• A relaxed, respectful manner: The interview tone—“a very relaxed atmosphere”—reflects the company’s broader culture, enabling open feedback and real exchange.
• Formal collaboration framework: The matrix org is the scaffolding that makes cross-team projects the default, not a special case.
• Strong technical focus: End-to-end product responsibility in hardware; clean handoffs from infra and backend to data science; intentional delivery to frontend in browser and apps.

It adds up to an environment where technical decisions interlock and teams move together from sensor to screen.

What candidates should bring: Social acuity, technical depth, and a love for interfaces

The session conveys a clear hiring profile:

• Authenticity: “Show yourself as you really are.”
• Social and technical strength: Both are actively explored in conversation.
• Cross-functional appetite: Projects involve multiple teams. Enjoy working at the boundaries and bridging disciplines.
• Quality mindset: Production and quality monitoring are part of the technical work, especially on the hardware/firmware side.
• Domain interest: Onboarding into dairy farming highlights how relevant context is to product decisions.

Why smaXtec stands out for engineers—tangible reasons

From the session, several differentiators emerge for smaXtec animal care GmbH:

• End-to-end ownership: A clear path from sensing to visualization, with teams accountable at every handoff.
• Modern pipeline: Wireless transmission, MQTT, Kafka, stream processing, and ML/NN—substantial challenges for those who enjoy real-time, data-driven systems.
• Matrix as the default: Continuous collaboration across disciplines fosters learning and shared problem-solving.
• Deep onboarding: Buddy system, department overviews, domain training, and close technical mentoring—conditions for confident ramp-up.
• Product, production, quality: A rare combination for hardware/firmware roles—product development tightly linked to production oversight and QA.
• Value-driven selection: A culture that values a relaxed, authentic conversation to find the right long-term fit.

Learning in the matrix: How growth happens here

The structure enables growth paths that feel natural rather than forced:

• Full lifecycle perspective: Firmware and electronics, data transmission, cloud ingestion, data science processing, and frontend delivery.
• First-hand feedback: Cross-team projects mean getting direct input from neighboring disciplines—e.g., how firmware choices affect stream processing, or how data science outputs translate to UI.
• Mentored ramp-up: The buddy system and close early support make it easier to contribute meaningfully while learning.

For engineers who want to broaden their scope—say, a firmware engineer curious about cloud ingestion, or a backend developer keen to better understand data science outputs—this structure provides daily opportunities.

Principles implied by the stack: Design and operations that reinforce each other

Even at a high level, the sensor → reading device → MQTT → cloud → Kafka → stream processing → ML/NN → frontend chain reveals practical engineering principles:

• Robustness at the edge: Data quality and firmware logic determine what’s feasible later in the cloud.
• Asynchronous decoupling: MQTT and Kafka create handoff points that decouple systems and aid scaling.
• Data as a product: Data science turns raw data into “interesting metrics.” Frontend usability is the ultimate validation of that transformation.
• Feedback loops: Insights from data science and frontend should inform firmware, backend, and infrastructure decisions.

These aren’t extras—they’re the backbone of a production system that blends physical sensing with streaming analytics and user-facing applications.

Interview structure mirrors collaboration: What the process signals

Having both HR and the Technical Team Lead in interviews is more than courtesy. It reflects how decisions are made:

• Clarity on the role: Expectations are aligned upfront.
• Shared evaluation: Social fit and technical competence are weighed together.
• Cultural preview: The “relaxed atmosphere” in interviews mirrors day-to-day collaboration.

Candidates effectively experience a slice of the working style during the selection process.

Domain context from day one: Technology in service of real use

“A small training is a direct entry into dairy farming,” says Frech. That matters. Domain understanding isn’t an optional add-on; it’s part of doing the job well. Knowing the context helps engineers make better decisions—from firmware sampling strategies to stream processing approaches to frontend signaling and visualizations.

For many engineers, that’s a rare opportunity: to deepen not just technical skills but also domain knowledge, and to build better systems because of it.

Quality as a throughline: From hardware to visualization

Because the hardware team also handles production and quality monitoring, quality becomes a continuous, end-to-end concern. It’s not an afterthought. Combined with the cloud and data pipeline, this setup encourages reliability at every stage—from the sensor onward.

Closing takeaways: A place for engineers who want to see their impact

From “Manuel Frech, Teamlead Hard- & Firmware bei smaXtec” (smaXtec animal care GmbH), we walk away with a coherent picture: clear functional teams, cross-functional projects by default, a modern data pipeline, a structured yet human recruiting approach, and onboarding that helps people grow into both the technology and the domain.

If you want end-to-end ownership—from microcontroller logic to metrics visible in a browser and on iOS/Android—this environment offers exactly that. It’s a place for engineers who value responsibility, love working at interfaces, and want to see the results of their work in a product that connects physical sensing with streaming analytics and practical, user-facing insights.

• 

  

  Johannes Pesenhofer, Data Scientist bei smaXtec

  Johannes Pesenhofer von smaXtec spricht im Interview darüber, wie er zu seiner aktuellen Arbeit als Data Scientist gekommen ist und was seiner Ansicht nach wichtig für Beginner ist.

  Watch now
• 

  

  Matthias Thym, Back End & System Engineer bei smaXtec

  Matthias Thym von smaXtec gibt im Interview Einblicke in seinen Background als Back End & System Engineer, wie seine aktuelle Arbeit abläuft und was für Anfänger in diesem Berufsfeld wichtig ist.

  Watch now
• 

  

  Oliver Troger, Front End Developer bei smaXtec

  Oliver Troger von smaXtec erzählt im Interview über seinen Werdegang, inklusive Umwegen bis zum Front End Development und gibt Tipps für Neueinsteiger.

  Watch now