dig leads the category of Social Video Intelligence, giving global brands the power to see, understand, and control the life of their brand on Social Video.

Our Algorithm Team doesn’t just build prototypeswe own the entire lifecycle from early-stage hypothesis and prompt/model architecture to high-throughput production deployment. We are looking for an Algorithm Engineer who blends the rigorous evaluation mindset of a data scientist with the execution speed and software craftsmanship of a production engineer.

About the Role

In this role, you will take ownership of two core pillars:

1. High-Throughput Multimodal Enrichment: Running large-scale LLM processing pipelines that ingest, parse, classify, and enrich millions of video posts per day with low latency and predictable unit economics.
2. Autonomous Research Agents: Designing and evaluating agentic workflows that autonomously reason over social video dynamics, analyze visual and textual narratives, and surface deep intelligence for our users.

You will bridge the gap between cutting-edge AI capabilities and production-grade reliability, placing massive emphasis on evaluation, dataset curation, alignment, and quality benchmarking.

Key Responsibilities

• End-to-End Ownership: Take agentic architectures and LLM/multimodal workflows from research, prompt engineering, and fine-tuning all the way to production microservices.
• Rigorous Evaluation Frameworks: Design, implement, and automate offline and online evaluation suites (e.g., LLM-as-a-judge, golden evaluation sets, human-in-the-loop validation, semantic consistency metrics) to quantify model output quality, hallucination rates, and drift.
• Large-Scale Data Enrichment: Build and optimize batch and streaming inference flows processing millions of daily video posts, balancing token budgets, concurrency, rate limits, latency, and accuracy.
• Agentic Workflow Development: Architect multi-agent systems, tool-use loops, retrieval-augmented generation (RAG), and reasoning traces tailored for social video analytics and market intelligence.
• Data Curation & Feedback Loops: Develop pipelines to collect implicit and explicit user feedback, curate fine-tuning datasets, and continuously refine model behavior.

Requirements

• Hands-on Experience: 3+ years of professional experience as an Algorithm Engineer, Data Scientist, or Machine Learning Engineer with a proven track record of shipping models to production.
• Applied LLM / Agent Background: Proven, hands-on industry experience building LLM-driven applications and/or developing research agents.
• Evaluation-First Mindset: Strong data science foundations with experience defining clear, objective benchmarks for subjective AI tasks (custom grading pipelines, error analysis, statistical significance testing).
• Software Engineering Skills: Advanced Python proficiency with high standards for clean, modular, test-driven code.
• Academic Background: B.Sc. in Computer Science, Data Science, Electrical Engineering, Mathematics, or a related quantitative field.

Advantages

• Agentic Frameworks: Hands-on experience with orchestration and agentic frameworks such as LangChain and LangGraph,
• Advanced Degree: M.Sc. or Ph.D. in Machine Learning, Data Science, Computer Science, or a closely related field.

What We Offer

• Direct ownership of critical, high-impact features powering millions of daily inferences.
• A flat, fast-moving engineering culture where research goes directly into production without bureaucratic handoffs.
• Access to state-of-the-art compute resources, modern tooling, and challenging multimodal datasets.