Sift

Plantd built a fully electric production line that turns fast-growing grass into carbon-negative structural panels. The largest homebuilder in America signed for 10 million of them. There was no service manual or industry baseline for normal. Every production run had to become part of the record. Once Plantd shifted from testing to manufacturing, engineers were manually reconciling machine telemetry, operator notes, and panel quality across separate systems. Now the whole loop lives on one timeline in Sift. Each run is compared automatically against the baseline built during commissioning, linking product quality back to machine behavior. Plantd spent late 2025 learning its new machine. That record made 2026 a production ramp year. Full story in the comments.

Telemetry at the edge should be this easy. See it in action on August 19. Registration link in the comments.

The Exploration Company is building the propulsion systems behind Europe’s return to orbit and its next push toward the Moon. With engineers across 5+ countries, TEC has to turn every hot fire into a clear next decision without slowing the program or compromising export control. Sift gives the team one platform to compare test runs, build calculated signals, and identify steady-state operation while keeping data hosted in France. The result is hours of manual analysis saved per test campaign. More time improving the engines. Less time maintaining the pipeline. Read the customer story in the comments.

Our mission is to advance science and technology by giving engineers the tools and infrastructure they need to build the future. Many companies at the forefront of that work run on Sift. We’re proud to work with K2 Space Corporation, Impulse Space, The Exploration Company, Astranis Space Technologies and many more. We refreshed our homepage to put their work front and center. Engineering teams move between testing, data review, root cause analysis, reporting, certification, and operations. When that work is scattered across a patchwork of software, teams lose context and time. Bringing it into one tool helps engineers move faster from prototype to ops while carrying what they learn from one stage to the next and across the whole organization. Homepage link in the comments.

Progress in hardware gets proven far from the office. A test out in the Mojave, a site with no network at all, a safety-critical operation where operators can't wait on the cloud. When the connection is weakest, the data matters most. Sift 𝐄𝐝𝐠𝐞, now available, runs full-fidelity capture right at the hardware. It ingests live sensor streams on a single local machine no matter how remote the site, writes every measurement to disk the moment it arrives, and serves live and historical views in a responsive desktop app. If the link drops mid-test, nothing is lost. Once connectivity returns, your team selects exactly which runs sync to Sift. Your existing integrations carry over. Point your Sift client libraries at Edge or stream over Arrow Flight, and everything your team already built on Sift runs unchanged. We're walking through all of it live in a webinar on 𝐌𝐨𝐧𝐝𝐚𝐲, 𝐀𝐮𝐠𝐮𝐬𝐭 𝟏𝟗 𝐚𝐭 𝟏𝟎:𝟎𝟎 𝐀𝐌 𝐏𝐓. Bring your questions. Learn more and register for the webinar in the comments.

Sift 𝗘𝗱𝗴𝗲 is here. The teams we build for are putting spacecraft into orbit, fielding autonomous systems, and defending nations. Their most valuable data is generated beyond the reach of any network, where it's hardest to watch live and easiest to lose. Until now, teams built a second stack of tools just to survive outages. Edge replaces that stack with one application. Telemetry is captured and viewable the moment it arrives, connection or not. After the test, teams choose exactly what syncs to Sift. The hardware our customers build has to work the first time. The proof now comes back from any test site, no matter how remote. Sift Edge is available on request. Learn more in the comments.

Parallel is moving short-haul freight off highways and onto rail with autonomous, battery-electric vehicles. Its in-house data stack worked when the company had ten engineers and one vehicle. As the fleet grew, storage costs climbed, analysis slowed, and engineers were pulled away from the vehicles just to keep the pipeline running. Parallel moved the entire stack to Sift. Infrastructure costs dropped 85%. The company now saves more than $140,000 a year in engineering time. Data review takes 150 fewer hours annually, and analysis that once slowed the team down now takes minutes per vehicle. Full story in the comments.

Hardware teams testing thrusters, flight vehicles, and reentry capsules all ask the same question after every test: Did this run behave like the ones before it? Answering this shouldn't mean hours across scripts, CSV exports, and one-off analysis tools. Sift Families does that comparison automatically. Pick your best runs as the baseline, and every new test gets reviewed against it the moment data lands. Reviews that took an afternoon now happen in seconds. Engineers spend their expertise on the deviations that matter, not rebuilding overlays. And when test cadence doubles, your team still gets faster. Read more in the comments.

Reliable Robotics Corporation is bringing autonomous flight to commercial aviation and defense. Its aircraft handle every phase of flight, from taxi to touchdown, while a remote pilot supervises from the ground. The FAA has accepted Reliable’s certification plan, the U.S. Air Force has the company under contract, and NASA is a collaborator. Reliable also completed the first FAA-approved uncrewed flight of a large cargo aircraft in U.S. history. Regulators need proof of how the aircraft performed in each test, what conditions it encountered, what inputs the system received, and how it responded. Sift brings the data from every flight, test, and simulation into one place, so engineering teams can trace back to the system actions, and test results behind them. Read more in the comments.

Astrolab is building a fleet of crewed lunar rovers for NASA - National Aeronautics and Space Administration’s Artemis program under a $219 million task order, with delivery targeted for 2028. On the Moon, there are no service calls. Every system has to be proven through thousands of hardware-in-the-loop, simulation, and field tests. That testing generates enormous volumes of telemetry. Most programs lose engineering time building and maintaining the infrastructure required to manage it, and the burden only grows as the program scales. Astrolab made the decision early to put every stream of telemetry on Sift and keep its engineers focused on the rover. The result: more than 5,000 engineering hours saved and redirected to the vehicle NASA is paying them to deliver. Learn more in the comments.