About WingRepWingRep is a fast-growing B2B SaaS startup redefining how sales teams level up. Our AI-powered platform delivers real-time sales coaching to accelerate revenue growthhelping sellers close more deals, managers scale their impact, and organizations drive consistent performance.Were building the next generation of sales enablement: smarter, faster, and more effective than traditional training. Our mission is to give every seller a WingRep coach in their corner.The team works out of our San Francisco office 3+ days a week, and we expect you to join us in the office at least 3 days a week to collaborate closely.The RoleWere hiring an experienced AI Engineer with strong backend chops to lead the development of our chat product. Youll own the systems that make WingRep feel magical: fast (<200ms) responses, contextually accurate coaching, and scalable AI infrastructure.This isnt just prompt hackingyoull be building the backbone of a next-gen AI platform. From async workflows and fine-tuning pipelines, to caching layers and RAG systems, youll be obsessed with speed, quality, and reliability.If the idea of building the worlds best sales coach excites you, we want you on the team.What Youll DoLead the technical direction of our chat product, while maintaining sub-200ms response times and world-class accuracy.Build and optimize LLM-powered systems, including fine-tuning, evaluation, and retrieval-augmented generation (RAG).Design and maintain async workflows for preprocessing meeting data, conversation memory, and tool orchestration.Architect and implement a caching layer (Amazon ElastiCache) to balance performance and cost.Work with AWS Bedrock (Claude, Llama), Pinecone, and Aurora Serverless to deliver highly contextual coaching.Collaborate closely with product and coaching experts to translate business goals into technical execution.Set engineering standards for speed, quality, and scalability across AI systems.Tech Stack Youll Work WithLanguages: PythonFrameworks/Services: FastAPI (running in Fargate), AWS Lambda, GraphQL APIAI/ML: AWS Bedrock (Claude, Llama), fine-tuning, embeddings, Pinecone RAGDatabases: Aurora Serverless (Postgres), ElastiCache (Redis)Infra: AWS (Fargate, S3, Lambda, etc.)Data: Heavy preprocessing pipelines for meetings, transcripts, and sales contextAbout YouYoure a backend-first engineer with 5+ years of experienceYouve worked with async workflows and know how to optimize pipelines for speed.You have experience with fine-tuning and evaluating LLMs (custom datasets, domain-specific tasks).You obsess over latency (<200ms) and quality of responsescutting corners frustrates you.Youre comfortable with distributed systems and caching strategies.You thrive in startup environments where speed of iteration matters as much as code quality.Bonus: Youve worked on sales tech, meeting assistants, or conversational AI.Why Join UsMission-driven: Were democratizing elite coaching, starting with saleswhere the impact is immediate and massive.High-impact role: Youll take point on building the worlds best AI sales coach.Startup velocity: Small team, fast iteration, direct influence on product direction.Personal growth: Work at the bleeding edge of AI infrastructure and applied LLM systems.Team culture: Were builders, runners, and doers (literallyour team ran the Livermore Half Marathon together in March 2025).Interview ProcessWe aim to move quickly while giving you the chance to showcase both your technical depth and product mindset:Phone Screen (30 min): Introductory conversation about your background, interests, and what excites you about WingRep.Systems & LLM Architecture Interview (6075 min): Deep dive into backend systems, async workflows, and applied LLM design.Pair programming: A focused challenge working with us to solve a touch challenge live.Onsite / In-Person Interview: Meet the team, review your take-home, and collaborate on a whiteboard session around scaling WingReps chat product.
recblid a9kbdq9isusaqz3w252niqzim9lfrp