Interview: Lead Engineer Behind the Open-Source Space Telemetry SDK

We recently spoke with Lina Kovács, the lead engineer of a community-maintained telemetry SDK that’s gaining traction among small-satellite teams and ground-station operators. Lina’s work straddles embedded systems, networking, and modern API design — valuable for both space and web developers building reliable telemetry pipelines.

Background and motivation

Programa.Space: Lina, what motivated you to start the SDK?

Lina Kovács: “We saw too many one-off telemetry stacks: custom formats, fragile encoders, and no clear upgrade path. Small teams need something reliable, lightweight, and extensible so they can focus on payloads and mission logic.”

Her background includes years working on embedded controllers and later on cloud infrastructure. That combination informed the SDK’s design goals: minimal runtime footprint, clear wire formats, and compatibility with modern cloud observability.

Design choices

Lina explained several pragmatic decisions:

• Binary but simple format: A compact binary envelope with TLV (type-length-value) encoding keeps bandwidth usage low while remaining extensible.
• Optional compression: For higher-latency links, optional delta encoding and lightweight compression reduce bytes sent without adding brittle complexity.
• Reconnection semantics: Built-in strategies for intermittent links, including graceful resync and journaling to local flash.

Interoperability with web stacks

One of the SDK’s strengths is how easily it plugs into modern backend stacks. An adapter layer converts telemetry messages into structured events compatible with Kafka, MQTT, or HTTP ingestion endpoints. This makes it straightforward to connect telemetry to dashboards or trigger server-side analysis pipelines.

Security and provenance

Security is often an afterthought in many small projects. Lina prioritized:

• Message signing: A small cryptographic signature that verifies origin without requiring heavy PKI.
• Replay protection: Sequence numbers and optional nonce-based schemes to prevent replay attacks.
• Audit trails: Metadata to track which software version and build produced a given telemetry stream.

Developer ergonomics

The SDK includes polyglot bindings — C for embedded, Rust for performance-sensitive modules, and Python/Node for rapid prototyping. Lina emphasized simplicity: “We provide batteries-included examples and a simulator so teams can iterate without hardware.”

Community and governance

The SDK is governed by a small steering committee with an emphasis on open contribution. Lina encouraged early adopters to submit real-world use cases and test data: “The more varied the missions, the more robust the tool becomes.”

Practical advice

Lina offered pragmatic tips for teams implementing telemetry:

• Start with a minimal set of fields and evolve the format rather than trying to anticipate every future need.
• Invest in a simulator — it speeds iteration by orders of magnitude.
• Design for graceful degradation: store locally and forward when connectivity returns.

Use cases and adopters

The SDK is used across several small-sat missions, hobbyist ground stations, and university labs. Lina highlighted a recent university project that used the telemetry stream to synchronize distributed sensor networks, reducing calibration time by 40%.

Future roadmap

Planned features include better tooling for schema evolution, a cloud-native ingestion service, and tighter integration with the Programa.Space monitoring stack. Lina also wants to support hardware-backed signing for higher-assurance missions.

Closing thoughts

The interview surfaces a recurring theme: pragmatic engineering wins. By balancing compact design, security basics, and simple integrations, the project reduces friction for teams launching real missions. If you’re working on telemetry, consider trying the SDK and contributing test vectors; the project benefits from diverse real-world datasets.

Want to read more? We’ll publish a follow-up that walks through integrating the SDK with a serverless ingestion pipeline on Programa.Space — including a sample simulator and dashboard configuration.