What Israeli startups actually build with — programming languages, open-source footprint, cloud-native and infrastructure maturity, and hiring stacks. Proprietary engineering signals from the startupim radar, read against external developer-survey and cloud-native data.
This report measures the technology choices of the Israeli startup ecosystem from signals companies emit — public source code, TLS-certificate footprints and live job postings — and reads them against external developer-survey and cloud-native data to separate what is distinctively Israeli from what is global.
Five findings stand out. (1) The language core is bilingual — JavaScript/TypeScript and Python dominate, mirroring the global pivot in which TypeScript became the #1 language on GitHub in 2025 (GitHub Octoverse). (2) Systems languages have arrived: Go appears in 47% of code-active companies and Rust in 15%, concentrated in cyber and infrastructure. (3) Israel's open-source weight is disproportionate but concentrated — a handful of globally-known projects (Redis, Aqua, Wix, Novu) anchor 1.66M stars, and the OSS culture is rooted heavily in the cyber / Unit 8200 cluster. (4) Infrastructure is maturing in step with the global cloud-native wave — Kubernetes is now in production for 82% of container users (CNCF 2025). (5) The teams writing this code are near-universally AI-augmented — yet 2025 saw Israel's first-ever decline in R&D employment, the central tension of the year.
Across the 1,195 companies with a matched public GitHub organisation, we counted the languages present in each org's repositories. Markup and config languages (HTML, CSS, Shell, Dockerfile) are excluded as non-application code.
sources.github.languages, June 2026. Share of the 1,195 companies with a matched public org. Companies use several languages, so columns do not sum to 100%.The shape is modern-web plus data/AI, and it tracks the global signal closely. JavaScript (40%) and Python (38%) run neck-and-neck; folding in TypeScript (27%) makes the JS family the most common technology in the ecosystem — consistent with GitHub's finding that TypeScript overtook Python and JavaScript to become the most-used language on GitHub in 2025, and with the Stack Overflow 2025 survey, where Python usage rose 7 points year-on-year and Rust and Go each gained 2 (Stack Overflow). The consequential local shift is the rise of Go (19%) past Ruby and C++, and the emergence of Rust (6%) as a real cohort.
Joining the code signal to each company's primary sector shows the "default stack" is really several stacks. The table shows, within each sector's code-active cohort, the share of companies using Python and Go.
| Sector | Code-active | Python | Go | Signature |
|---|---|---|---|---|
| Business / Enterprise Software | 440 | 40% | 23% | TypeScript-forward SaaS |
| Cyber Security | 196 | 52% | 30% | Go + Python, Rust-curious |
| Health Tech & Life Sciences | 119 | 19% | 10% | Mixed, app-layer JS |
| Fintech & Insurtech | 103 | 30% | 14% | JVM + TypeScript |
| Media & Entertainment Tech | 81 | 28% | 12% | TypeScript-led |
| Industrial Technologies | 75 | 48% | 8% | Python + C/C++ embedded |
| Automotive & Mobility | 44 | 30% | 18% | C++/Python + Go services |
startupim.companies.primary_sector_name, June 2026. "Code-active" = sector companies with a matched public GitHub org.Cyber Security is the standout — the most Go-heavy sector (30%), the most Python-dense (52%), and home to the largest Rust cohort: the signature of a sector building high-performance agents, scanners and network tooling. Industrial and Automotive retain a C/C++ embedded core with Python for data and modelling, while Fintech keeps a conservative JVM lean alongside TypeScript front ends.
Israeli companies punch far above their number in open source. The public repositories of matched orgs carry 1.66 million GitHub stars in aggregate, anchored by a small set of globally-recognised projects.
sources.github.total_stars / top_repos, June 2026. Stars summed across each matched org's public repos. Illustrative of the OSS-anchor tail, not a company ranking.The pattern is strategic: Israeli infrastructure and dev-tools companies use open source as a distribution and credibility
channel — Redis, Aqua's trivy, Appwrite, Novu and DragonflyDB are each category-defining projects that seed a
commercial business. The lineage runs deep: Elasticsearch (co-founded by Israel's Shay Banon), JFrog,
Snyk and others, among others, helped make Israel a genuine commercial-OSS hub. But there is a sceptic's counterpoint worth stating: much of
this strength is concentrated in cybersecurity and the Unit 8200 alumni network — a tight, relationship-based cluster
that some argue is narrower than a "broad ecosystem." Both readings are visible in our data: the stars are real, and they
cluster in infra/cyber.
last_push fields make it possible to spot the next anchor early; expect it to emerge from infra or security.Public TLS certificates leak the shape of a company's infrastructure. Of 4,159 companies with usable certificate-transparency data, we bucketed sub-domains to gauge dedicated API, application, staging and status environments.
sources.crtsh.buckets (crt.sh certificate transparency), June 2026. A bucket counts a company once if any matching sub-domain has appeared in a logged certificate.An api. host (14%) and an app. host (13%) mark a product-shipping company with a separated
service tier; the ~1-in-10 rate of public staging./dev. environments indicates a maturing CI/CD
culture. This local picture sits inside a decisive global wave: the CNCF's 2025 survey reports 82% of container users run
Kubernetes in production (up from 66% in 2023), 98% have adopted cloud-native techniques, and — tellingly for an
AI-heavy ecosystem — 66% of organisations hosting generative-AI models use Kubernetes for inference
(CNCF).
api./staging./status. trio separates a company shipping a product on real
infrastructure from one running a single marketing site — useful for diligence, sales targeting and competitive mapping.For the 322 companies where a live applicant-tracking system was detected, the choice of ATS is itself a stack signal — and it has consolidated onto modern, developer-first platforms.
sources.jobs.ats, June 2026. Detected via careers-page scrape or direct ATS probe. ¹Other = SmartRecruiters (6), BambooHR (5), Breezy (3), Comeet (1).Greenhouse and Ashby account for two-thirds of detected hiring stacks. Ashby's strong second place (30%) is a generational marker — the newer, analytics-first system favoured by AI-era startups. Notably, the homegrown Israeli ATS Comeet appears only once; local startups overwhelmingly reach for global platforms.
But the most important hiring story of 2026 is not the tool — it is who is being hired, and whether. The teams these companies run are now near-universally AI-augmented: a joint Israel Innovation Authority / Startup Nation Central survey finds 95% of high-tech employees use AI tools regularly, 78% daily, with Israel among the highest-ranked countries worldwide in Claude usage per working-age person, and labour productivity in high-tech up 4.7% (CTech). Yet the same year brought Israel's first-ever decline in high-tech R&D employment (−1.1%) — the reconciling thesis being that AI lets companies ship more with fewer engineers.
The global survey data adds a cautionary note: developer AI-tool adoption hit 80%, but trust in AI accuracy fell from 40% to 29%, and 66% of developers say they spend more time fixing "almost-right" AI-generated code (Stack Overflow 2025) — a productivity-versus-quality tension now baked into every engineering org.
Three trajectories follow. First, Python keeps rising on the AI build-out, and the JS/TS family stays the default product layer. Second, systems languages consolidate — Go is table-stakes for infrastructure and cyber, and Rust's cyber-led cohort is the one to watch. Third, infrastructure maturity becomes a moat signal — the separation of API, staging and status environments, visible for free in certificate logs, increasingly distinguishes enterprise-ready companies, against a backdrop where Kubernetes is now the de-facto substrate for AI workloads.
The watch-item underneath all three is talent. Israel holds the world's highest concentration of AI talent (LinkedIn AI Talent Index, 1.98% of the workforce), but 2025's R&D-employment decline — and reports of startups building dev teams abroad for lower cost — raise a question the report's author for the State of High-Tech 2025 put plainly: policymakers must "monitor the trend closely to ensure it does not signal the beginning of a brain drain" (CTech). The stack is healthy; the question is where, and by how many, it will be built.