TenderStria is building the AI that reads public-sector tenders so teams don't have to. Every working day, governments across 50+ countries publish thousands of contract notices, far more than any bid team can read. We ingest them all, translate and normalise them, and score each one for fit against a company's services, sectors and bid criteria, so teams open only the contracts genuinely worth pursuing. We're an early, deliberately small team building toward our first customers, with a product that already works.
Every day, thousands of tenders arrive in dozens of languages and formats, and we have to read, translate, normalise and score every one, accurately, cheaply, and fast enough to alert teams the same day. You'll own significant parts of how we turn that flood of documents into accurate, scored results, and the quality bar that keeps it honest as it grows.
This is a role for someone who cares about quality you can measure. A wrong score, a missed requirement, a silently dropped tender, these cost customers trust. You'll be expected to catch them before they ship, and to make the system better and cheaper over time.
Most 'AI engineer' roles are a thin wrapper over someone else's API. Here the system is the product, and its quality is the moat. You'll own real surface area, measure your impact directly, and work on extraction and scoring problems that genuinely decide whether a customer wins work.
Public procurement is a vast market that still runs on tools barely better than email alerts. We think the team that makes it genuinely searchable and scoreable will define the category. You'd join early enough to shape the product, the culture, and the playbook, with the ownership that only an early team offers and the focus of a company that has chosen its problem deliberately.