One Role, Remote-First
Orbio Earth, a Y Combinator Summer 2023 graduate, is recruiting a founding product engineer for its methane monitoring platform. The Y Combinator Work at a Startup listing describes the role as an "exceptional engineer to own and scale our methane monitoring product," working directly with the CEO on end-to-end responsibility for core systems: maintaining and improving shipping workflows, automating QA and reviewing systems, launching a new front end, maintaining geospatial pipelines, building APIs and integrations for operators, and optimizing infrastructure costs. The position is listed as remote; CB Insights identifies the company's base as Cologne, Germany.
Orbio builds a platform that ingests satellite imagery (primarily freely available Sentinel-2 data, with the company "working to apply its algorithms to other satellite data"), runs detection models, and sells quantified emissions data to financial institutions and energy companies. TechCrunch reported that Orbio processes 10–20 terabytes of data daily and delivers emissions data to end users within 48 hours. The company's LinkedIn page states its mission as "accelerating the low-carbon transition with actionable methane intelligence."
Project Eucalyptus, Orbio's open technical tutorial published on GitHub, documents how the team stitches together multi-sensor inputs to overcome revisit-rate limits and cloud cover. The founding product engineer inherits ownership of this chain: ingestion, inference serving, and the API that delivers flux estimates to downstream consumers.
Finance First, Operators Second
Orbio's customer mix differs from many climate-tech startups. Startupintros reports the company's model: "We sell the data to financial companies that want to invest into the best performing energy companies." TechCrunch noted that customers include both oil and gas operators monitoring their assets and investors assessing portfolio performance or benchmarking competitors.
The arithmetic is stark. The IEA states that methane has driven about 30% of climate change. Traditional emission-factor methods underestimate actual emissions by up to 70%, per IEA data cited on Orbio's website. The U.S. EPA methane fee is $900 per metric ton in 2024, rising to $1,500 by 2026. Orbio estimates that onshore U.S. oil and gas exploration and production released about 10 million metric tons of methane in 2023; at the 2026 rate, that exposure reaches $15 billion annually — a liability that ultimately hits equity and debt holders.
Orbio's platform replaces emission factors with satellite-verified, asset-level data stretching back to 2016, letting investors benchmark operators, price risk, and flag abatement opportunities before regulators do. The 48-hour delivery SLA serves trading-desk workflows; the API feeds slot into quant models. The founding product engineer will own the pipeline that turns daily terabytes of Sentinel-2 imagery into auditable, financially material emissions estimates, with a team of six (per Reforgers data), reporting to founders Robert Huppertz and Jack Angela, who bring expertise from NASA Harvest and Cervest.
The Capital Clock
Orbio graduated from YC S23, a cohort that included several climate-data ventures, and has raised $4.6 million across two rounds: a $600,000 pre-seed in September 2022 led by Romain Diaz, and a $4 million seed in March 2024 from the European Space Agency, Initialized Capital, Y Combinator, David Rusenko, JJ Fliegelman, and Liz Wessel (Startupintros). The founding-engineer search now underway carries the hallmarks of a company preparing for Series A: a single hire to own the full stack (satellite ingestion, ML detection, quantification, and delivery) before headcount expands into specialized squads.
YC-backed climate data startups that reached Series A fastest typically locked a design partner in a regulated market (insurance, commodity trading, or sovereign MRV) and built the API layer that partner required. Orbio's choice to lead with financial institutions suggests it is following that playbook. The founding engineer will likely spend the next 12–18 months hardening that API, standardizing the quantification methodology for auditability, and documenting the error bounds that a counterparty's risk model can ingest.
Where Orbio Stands
Orbio enters a field where the technical bar has been set by two better-capitalized incumbents, GHGSat and Kayrros, and a new wave of purpose-built constellations. The hiring profile reflects that follows from that reality: the role must bridge orbital-mechanics literacy, ML detection pipelines, and quant-finance fluency because the product sits atop data that competitors have spent a decade validating.
Three Technical Paths
| Company | Founded | Core architecture | Detection limit (validated) | Primary customer | Capital position |
|---|---|---|---|---|---|
| GHGSat | 2011 | Own constellation (C1–C6+) | 0.20 t/h (±13%) | Operators, regulators | $100M+ raised; commercial revenue |
| Kayrros | 2016 | Multi-sat fusion + AI (Sentinel-5P/2, PRISMA, WV3, GHGSat) | 1.4 t/h (Sentinel-2) to 0.2 t/h (GHGSat tasking) | Governments, IEA, UNEP, asset managers, operators | Acquired by Energy Aspects (May 2026) |
| Orbio Earth | 2021 | Satellite-agnostic platform (all public + commercial imagery) | Not yet single-blind validated | Financial institutions, energy portfolios | YC S23; $4.6M raised; pre-Series A |
Sources: CB Insights (GHGSat founding, location, customer focus); Nature single-blind study (detection limits, quantification accuracy); Kayrros website (customers, acquisition); TechCrunch, Startupintros, Orbio website (Orbio founding, architecture, customers, funding).
The Nature single-blind validation — the only independent, controlled-release test of its kind — showed GHGSat-C2 detecting 0.20 t/h releases with stage-1 quantification error between −17% and +13%, the tightest of any satellite-team pair tested. Kayrros operates as an analysis layer: it fuses Sentinel-5P TROPOMI (daily, >10 t/h), Sentinel-2 and Landsat 8 (10–16 day revisit, ~1.4 t/h), and tasked high-resolution assets (PRISMA, WorldView-3, GHGSat) into a single API. In the same study, Kayrros analysts participated as one of five independent teams; their results fell within the aggregate 75% of estimates inside ±50% of metered value.
Its platform ingests every available public overpass (Sentinel-2, Landsat, TROPOMI, and tasked commercial frames) and runs a unified detection-to-quantification pipeline calibrated against the same Stanford controlled-release dataset that validated GHGSat. Orbio's website cites a Stanford case study confirming "market-leading emissions data with global, asset-level satellite imagery."
Business Models Diverge
GHGSat sells measurements to operators and regulators needing facility-level compliance proof (CB Insights). Kayrros sells intelligence subscriptions to the IEA, UNEP's MARS program, the World Bank, and asset managers modeling portfolio methane risk (Kayrros Methane Watch). Orbio's website leads with: "Reduce energy portfolio risk with global methane emission tracking" — the IRA methane fee and EU Methane Regulation turn satellite data into a tradable risk factor. That difference shapes the engineering hire. GHGSat needs spectrometer physicists and flight-software engineers. Kayrros hires geospatial data scientists and energy-market analysts. Orbio needs a product engineer who can turn a heterogeneous, multi-satellite feed into a quant-grade emissions factor that a credit desk can ingest via API.
The Validation Hurdle
The Nature study established a public benchmark: 71% detection rate across 49 satellite-team measurements, zero false positives, quantification parity with airborne remote sensing (Kairos Aerospace: 86% within ±50%). GHGSat's purpose-built hardware won on sensitivity; Kayrros's fusion approach won on coverage. Orbio has not yet published single-blind results. Until it does, financial customers have no independent basis to prefer Orbio's asset-level factors over Kayrros's IEA-endorsed inventories or GHGSat's direct measurements.
Headcount and Capital
Reforgers reports Orbio's team size at six. GHGSat and Kayrros each employ dedicated squads for ingestion, detection, quantification, and API layers. Orbio's founding product engineer will write the ingestion layer, the plume-detection model, the quantification engine, and the API contract, solo or in a pair. The hiring bar isn't "can you build this?" It's "can you build this at a velocity that keeps the product ahead of the next Carbon Mapper or MethaneSAT data drop?"
Carbon Mapper's constellation (designed detection limit 0.05–0.15 t/h, 1–7 day revisit at full deployment) and MethaneSAT (EDF-backed, point-source plus wide-area flux) will flood the market with higher-cadence, lower-threshold data within 18 months (Nature study). Orbio's moat is the product layer that normalizes every sensor into a single, auditable emissions factor per asset per day. The founding engineer builds that normalization layer.
A New Hybrid Workforce
The BLS projects data scientist employment to grow 34% from 2024 to 2034 — roughly 23,400 openings per year — with a median wage of $112,590. A 2024 Kaggle dataset of 485 data-science postings shows the field fragmenting into specialized tracks: ML engineering, quant research, and geospatial analytics each now command distinct salary bands and skill stacks (GitHub: adeenaamersi/Navigating-the-Data-Science-Job-Market).
Satellite methane intelligence sits at the intersection of all three. The MethaneSAT program demonstrates the orbital-mechanics side: engineers who understand sensor tasking, revisit rates, and atmospheric radiative transfer. The quant-finance side appears in industry reporting on climate risk modelling and real-time analytics lengthening search cycles. Orbio's founding product engineer must speak all three languages: task a bird, train the detection model, and structure the output so a credit desk can price it.
UNjobs lists "methane emission data" as a sought skill, confirming institutional demand. The talent pipeline remains thin: traditional remote-sensing programs don't teach financial product design; quant programs don't teach orbital mechanics. The next five years will test whether universities and bootcamps can produce this hybrid at scale, or whether companies like Orbio continue competing for the same small pool.
Working in frontier tech? Zero G Talent tracks the openings: browse frontier tech jobs, openings at ASML and Stripe, and the people building the field.