• Work in a team to deliver state-of-the-art closed- and open- world tracking systems for L4 self-driving vehicles. • Contribute to the technical direction of the tracking team and broader perception org. • Write robust software for real-time, resource-constrained, safety-critical applications. • Drive highly cross-functional features and work streams from problem statement to delivery. • Proactively identify limitations with the existing perception system and drive improvements. • Root-cause issues that arise onboard, and provide solutions. • Align stakeholders with strong presentation and communication skills. • Lead challenging technical topics and build consensus.

• Bellese is looking for an AI-enabled product engineer to rapidly develop prototypes and proofs of concept. • You will lead a small team of junior technologists in building solutions to early-stage ideas formulated by various teams and via organic ideation sessions. • You will work to rapidly build working software that can demonstrate real business impact to clients. • You’ll be focused on shipping products in weeks, not months. • You won’t just determine system architecture, you’ll shape how products deliver tangible client value. • You’ll show resilience and creativity in iterating on and evaluating technology solutions posed; you won’t be afraid to move on from approaches that aren’t working. • This role reimagines what an engineer looks like. You will not be handed fully fledged product specifications or features, rather, you will need to take nascent ideas and turn them into working software. • You will own more of the product lifecycle than a traditional software engineer.

• We are looking for a highly skilled developer who is passionate about technology to join our development team.

• Design and build advanced Clay tables from scratch to support outbound and enrichment campaigns at scale • Create structured multi-step Clay workflows that handle thousands of records without breaking or requiring manual intervention • Build and maintain reusable Clay data models that support sales, marketing, and RevOps use cases simultaneously • Own the full Clay table lifecycle from initial architecture through ongoing optimization and maintenance • Use Claude as the primary intelligence layer for all Claygent prompt development and structured research automation inside Clay • Write, test, and iterate on Claude prompts until outputs are structured, reliable, and production-ready, not just good enough • Use Claude to extract meaningful signals from websites, LinkedIn profiles, job postings, and unstructured web sources • Continuously refine Claude prompt logic based on output quality, accuracy, and enrichment coverage • Optimize the balance between Claude model selection, prompt complexity, and Clay credit spend for maximum efficiency • Design and implement waterfall enrichment sequences across multiple Clay-connected data providers • Continuously test and improve provider sequencing to maximize coverage while minimizing wasted credits • Diagnose enrichment gaps, monitor accuracy across live Clay workflows, and implement fixes proactively • Integrate Clay outputs directly into Salesforce or HubSpot with clean field mapping, structured data syncing, and automated triggers • Ensure enriched Clay data flows into CRM pipelines without hygiene issues or data loss • Build outbound activation logic that connects Clay-enriched records to sequencing tools and campaign execution • Partner with sales and RevOps to translate targeting requirements into Clay and Claude-powered technical workflows • Support outbound campaigns running at 5,000 or more leads per month using Clay as the core data engine • Proactively surface optimization opportunities across the full Clay and Claude stack