• Research, implement, and validate model architecture and algorithm changes that improve video generation fidelity, with emphasis on human-centric quality (identity preservation, anatomy, motion coherence, and interaction realism) • Explore and prototype improvements across spatial multimodal modeling, modality alignment, flow-based or diffusion-based video generation, and neural rendering-inspired representations to improve controllability and long-horizon consistency • Improve training and inference efficiency through architectural and post-training techniques (compute/memory optimizations, distillation, pruning, and compression) • Define model training objectives that improve sim-to-real and real-to-sim generalization, especially for human motion, contact, and interaction dynamics across real-world and synthetic/simulation data • Develop detailed, domain-specific benchmarks for evaluating world foundation models, especially generation and understanding world models that reason about video, simulation, and physical environments • Translate research results into robust implementations like training code, production-grade checkpoints, model integrations, and demos that clearly showcase capability gains across teams

This description is a summary of our understanding of the job description. Click on 'Apply' button to find out more. Role Description As an Applied AI Scientist on our Support Automation product team, you'll: - Bridge research and production, building AI models and agent systems that ship into real products. - Architect, evaluate, and optimize autonomous voice and digital agents powering 10M+ customer interactions per month across West Africa. - Build for the hardest edge cases: poor connectivity, low literacy, and languages with little training data. - Experiment with, evaluate, and integrate the latest voice and text models. - Own problems end-to-end, from problem discovery to running in production, working alongside product and engineering leaders who prioritize shipping real customer impact. If you're energized by ownership, thrive on deep technical challenges, and want to build infrastructure that serves millions in emerging markets, let's talk. Our stack (prior experience is a strong plus, but not required): - Backend: Python 3 (+ mypy) - API layer: GraphQL - Android frontend: Kotlin/Jetpack - iOS frontend: Swift/SwiftUI - Web frontend: TypeScript/React - Database: Postgres/CockroachDB - Infrastructure: GCP/Terraform - Orchestration: Kubernetes Qualifications - 5+ years of experience in AI/ML engineering - Strong Python and backend engineering skills - Solid foundation in statistics and ML theory - Proven hands-on experience building with LLMs (prompt engineering, RAG, embeddings, fine-tuning, agent orchestration etc.) - Track record taking ML models from prototype to production and care deeply about reliability, performance, and scalability. - Fluency with AI coding agents, with bonus points for deploying them as part of production systems Requirements - Research background in speech or LLMs - Proven Voice AI experience (ASR, TTS, latency optimization, real-time streaming, voice agents) - Experience building customer support or conversational AI systems - You follow the latest research in speech and LLMs, and know how to translate papers into production wins - Familiarity with low-resource language challenges - Fluent in English (bilingual in French is a big bonus!) Benefits - This is a fully remote role. Candidates must be based in one of our talent hub countries (US, Canada, UK, Spain, Kenya and Ghana) or in one of our operating markets in Africa including Senegal, Côte d'Ivoire, or Burkina Faso. - Wave provides a yearly $1,200 stipend to support coworking meetups with teammates. - Remote team members are expected to travel to our operational markets (e.g. Senegal or Côte d'Ivoire) at least once a year. - Our salaries are competitive and are calculated using a transparent formula. For this role, depending on your level and location, we offer a salary up to $227,900, plus a generous equity package. - Subsidized health insurance for you and your dependents and retirement contributions (both vary from country to country). - 6 months of fully paid parental leave and subsidized fertility assistance. - Flexible vacation, with most folks taking between 21-30 days exclusive of statutory holidays. - $10,000 annual charitable donation matching.

This description is a summary of our understanding of the job description. Click on 'Apply' button to find out more. Role Description We’re building several LLM-powered copilots across critical workflows (e.g., underwriting productivity, agent enablement, customer support, operations/compliance, fraud). We need an AI engineer to own the LLM + retrieval + context layer that makes these copilots accurate, auditable, fast, and cost-efficient. Typical stack: Python/FastAPI, Postgres + vector (pgvector/Pinecone/Weaviate), OpenSearch, optional graph DB, Kubernetes + GPUs, OTEL/Datadog - Production RAG: indexing, retrieval, hybrid search, reranking, query rewriting, grounding, citations - Context Graph: entity resolution + linking + provenance; graph + vector retrieval; supports multi-hop context - LLM orchestration: tool/function calling, structured outputs, routing across model tiers, failure modes - GPU/inference cost optimization: batching, caching/KV reuse, quantization, autoscaling; optimize $/session + latency - Safety + compliance: PII/PHI handling, redaction, audit logs, deterministic replay, hallucination mitigation - LLMOps: eval harness (golden sets, regression, adversarial), monitoring for quality/cost/drift - Design/ship the end-to-end pipeline: retrieve → assemble context → generate → cite → log/monitor - Improve quality and trust via evaluation, feedback loops, and clear evidence-backed outputs - Partner with product, security, and domain teams; write crisp design docs; raise engineering bar - Ship RAG v1 with citations + measurable quality metrics - Deliver Context Graph v1 that improves retrieval on real copilot tasks - Reduce cost/latency with a concrete inference optimization plan shipped to prod Qualifications - 7+ years building production systems; 2+ years hands-on LLMs/RAG - Proven RAG experience (embeddings, vector DBs, hybrid search, reranking, eval) - Strong backend/distributed systems + observability - Track record shipping in high-stakes environments with auditability/correctness - Knowledge graph / entity resolution / provenance systems - GPU inference optimization (vLLM/TGI/TensorRT-LLM, quantization AWQ/GPTQ, batching) - Regulated domain experience (insurance/fintech/healthcare) Requirements - #LI-Remote - #LI-MK1 Benefits - The US national base salary range for this full-time position is $133,000 - $235,000. - Our salary ranges are determined by role, level, and location. - Individual pay is determined by work location and additional factors, including job-related skills, experience, and relevant education or training. - Compensation details listed in US role postings reflect the base salary only and do not include applicable bonus, equity, or benefits. - Further details of our US benefits can be found at Ethos Careers .

• Agentic Engineering & Orchestration Workflow Design: Architect complex, multi-agent workflows using Microsoft AI tech stack. • Design, Develop and Deploy agents to handle loops, interruptions, and human-in-the-loop interventions. • Tool Use & Function Calling: Build reliable "tool layers" that allow LLMs to safely interact with internal APIs, databases, and third-party SaaS platforms (e.g., Salesforce, Workday, ServiceNow etc.). • State Management: Design persistence layers to manage agent memory, conversational history, and context windows efficiently. • Advanced Data & RAG Strategy Retrieval Pipelines: Build production-grade data retrieval and integration systems. • Optimize vector indexing, document chunking, and re-ranking algorithms to ensure high-precision context retrieval. • Data Quality: Collaborate with Data Engineers to curate "Golden Datasets" for agent consumption. • LLMOps, Evaluation & Quality Automated Evaluation: Build CI/CD pipelines for AI that include "LLM-as-a-Judge" testing. • Leverage frameworks to score agent outputs for accuracy, hallucination, and safety before deployment. • Observability: Instrument applications with tracing tools to visualize agent reasoning chains, monitor latency, and debug failures in production. • Cost Optimization: Monitor token usage and latency, optimizing prompt density and caching strategies to maintain high performance at sustainable costs. • Innovation & Collaboration: Prototyping to Production: Rapidly validate new ideas using state-of-the-art models, then refactor successful prototypes into maintainable, tested production code. • Standards Adoption: Stay ahead of the curve by evaluating emerging technologies to standardize agent connectivity.