AllCode VS Talents-OS

Talents-OS's answer:

Talents-OS replaces the résumé with evidence. Every candidate completes a verification journey — auto-graded technical tests, an analysis of their real GitHub activity, and an adaptive AI interview — which combine into an explainable TalentOS score, so a company sees not just a ranking but why a candidate fits and where the gaps are.

Two things set it apart. First, it is Arabic-first and RTL-native, built for the Arab world across 22+ countries rather than translated into it, covering 90+ technical specializations. Second, it measures how a candidate relates to AI in both directions: an integrity layer detects assessments outsourced to an external model, while a separate evaluation track hands the candidate an AI-generated solution with real flaws planted in it and scores them on catching and correcting those flaws. Resisting AI misuse and working skilfully with AI are different skills, and both matter now.

Talents-OS's answer:

Most technical-screening tools answer "did this person pass a test?" Talents-OS answers "what can this person actually do, and how do we know?" The score is explainable and traceable to evidence rather than a single pass/fail number.

Three practical differences:

— Arabic-first, not localised. Assessments, interviews and reports work natively in Arabic with full English support. Global platforms treat Arabic as a translation layer, which does not survive a technical interview.

— Blind by default. The first evaluation is on competence alone; identity is revealed afterwards. That is built into the flow, not an optional setting.

— Built for AI-era assessment. Detecting an outsourced answer and measuring genuine skill at working with AI are treated as two separate measurements, because they are.

Talents-OS is early and does not pretend otherwise. Companies start with their first 2 job posts free, so the evidence can be judged directly rather than taken on our word. For pure algorithmic drill-testing at scale, HackerRank, Codility and CodeSignal are mature and good at it — they are simply solving a different problem in a different market.

Talents-OS's answer:

Two sides of one market.

Companies hiring technical talent in the Arab world — startups and engineering teams across the Gulf, the Levant, Egypt and the Maghreb who need to know whether a candidate can actually do the work before committing to interview loops, and who are hiring in Arabic-speaking markets where global screening tools fit awkwardly.

Developers, designers, data and security professionals in the region who can do the work but whose CV does not prove it — including self-taught engineers and career-changers without a recognised degree or a well-known employer on their record. For them, a verified score is a way to compete on demonstrated ability.

The platform covers 90+ specializations across software engineering, data and AI, design, cybersecurity and product.

Talents-OS's answer:

Talents-OS was founded in Amman, Jordan in 2026 around a specific mismatch: in the Arab world, hiring decisions still lean heavily on résumés and credentials, while the fastest-growing pool of capable technical people is self-taught and credential-light. Companies filter on signals that do not predict performance, and capable candidates are filtered out before anyone looks at their work.

The response was to build the evaluation layer instead of another job board: verify the skill directly, explain the result, and let companies hire on proof. Arabic-first was a deliberate architectural choice rather than a market afterthought — an assessment or an interview that only half works in a candidate's own language does not measure skill, it measures language.

The platform is early. It is being built in the open, and the honest state of it is published rather than dressed up.

Talents-OS's answer:

Talents-OS runs entirely on Cloudflare's edge platform.

— Frontend: React 19 with TypeScript and Vite, served from Cloudflare Pages, bilingual and RTL-native, with marketing pages pre-rendered for crawlability. — Backend: Cloudflare Workers exposing both REST and tRPC APIs. — Data: Cloudflare D1 (SQLite at the edge) with Drizzle ORM, R2 for object storage, and Durable Objects for stateful real-time features such as chat and interview rooms. — AI: Cloudflare Workers AI for assessment grading, matching and embeddings, with all model calls routed through a single instrumented layer. — Mobile: a full Flutter application sharing the same API contract as the web client, with a generated contract file keeping the two from drifting apart. — Supporting services: Stripe for payments, Resend for transactional email, Sentry for error tracking and PostHog for product analytics.

Deployment is fully automated: every merge to the main branch runs type-checking, the client and worker builds, and the full test suite before shipping.