full stack engineer

Jiayu Zhao

Chicago, IL 872-***-**** *********@********.*** Linkedin github.com/jiayuzhao05

EDUCATION

University of Chicago Chicago, IL

M.S. in Computational Analytics and Public Policy Sep 2024 – Jul 2026(Expected) Donghua University Shanghai, China

M.S. in Accounting Sep 2018 – Jul 2020

Hunan Normal University Changsha, China

B.S. in Accounting Sep 2013 – Jul 2017

PROFESSIONAL EXPERIENCE

DeepManifold Inc. New York, NY

Software Engineer Intern Mar 2026 – May 2026

● Built a real-time fraud and merchant risk pipeline on Kafka/Flink, integrating authorization, KYC, chargeback, and behavior signals with <50ms streaming latency.

● Productionized a Feature Store for transaction, behavior, merchant, and graph features, ensuring online/offline consistency with minute-level refresh for serving-time risk signals.

● Served low-latency scoring and decisioning via TensorFlow Serving + Redis, meeting a 100ms SLA; produced risk scores

(0–1) and routed actions (allow, step-up, block) with config-driven strategies supporting A/B tests and canary rollouts.

● Closed the feedback loop with chargeback and reviewer labels for daily incremental updates and weekly retraining; added SHAP-based explainability and monitoring (decision logs, Grafana/Prometheus) targeting Recall >92% and FP <1.5%. Happy Cashier Inc. Chicago, IL

Software Engineer Intern Jun 2025 – Aug 2025

● Built and shipped a restaurant POS merchant portal end-to-end (front-end + service integration), supporting hundreds of merchant accounts/month; hardened external calls with timeouts, retries, and graceful degradation to handle API failures.

● Integrated backend APIs and reduced average response latency from ~800ms to ~350ms via profiling and performance tuning

(request flow and payload optimization).

● Built an NLP-based support deflection module covering top intents (~90% of ticket volume) and added speech-to-text for 24/7 self-serve support, reducing escalations by ~60% post-launch. Boyu International Trading Co., Ltd. Remote

Software Engineer Intern Jul 2024 – Jan 2025

● Built and launched a full-stack e-commerce website for physical 3C accessories, serving real orders and onboarding 1,000+ registered users within 3 months.

● Integrated Stripe/PayPal checkout and processed $25K+ in transactions; handled payment states (success/failure/cancel) and recorded order–payment mappings to prevent mismatches.

● Developed backend services using Django + PostgreSQL, implementing core modules for product catalog, inventory, authentication, and order management with reliable persistence. Bank of Communications Shanghai, China

Quantitative Risk Analyst Sep 2020 – Aug 2023

● Built Python/SQL-based feature engineering and risk modeling workflows on customer, transaction, and repayment data, supporting credit decisioning through model monitoring, backtesting, and threshold optimization. PROJECT EXPERIENCE

Real-Time Bidding & Budget Optimization Engine (RTB + RL) Jun 2025 – Sep 2025

● Built an industrial-grade RTB decisioning engine with Flink/Kafka/Redis + feature store, sustaining 100K+ QPS and 20–50ms end-to-end latency for auction-time bidding at scale.

● Engineered real-time feature pipelines for user embeddings (DeepFM/DSSM) and CTR/CVR feedback loops, achieving sub-50ms feature availability before impression auctions.

● Implemented market-aware bidding and pacing by combining pCTR/pCVR, bid-landscape signals, and RL-driven strategies

(contextual bandits for cold-start exploration; deep RL for budget pacing/frequency control) under constraint guardrails.

● Delivered online impact across campaigns with monitoring and rollback-ready serving: +18.5% ROI, -10% budget consumption at equal GMV, +12% CTR lift (cold-start), and -35% early budget burn-out. AI-Enhanced News Aggregation & Engagement Platform (Full-Stack + ML Service) Oct 2024 – Jan 2025

● Built a production news aggregation platform with a push notification pipeline, supporting 20K+ active users and ensuring reliable delivery through robust retry/failure handling.

● Developed low-latency backend APIs (Express + MongoDB) for feed and personalization endpoints; achieved sub-200ms response latency via query tuning and service-level profiling.

● Implemented Redis cache-aside with TTL and invalidation to reduce data retrieval time by ~50% and improve responsiveness under load, especially for hot feeds and repeated access patterns.

● Integrated a FastAPI-based generative content enrichment service and productionized it on AWS (Lambda/ECS) with safe failure handling (timeouts, retries, fallbacks) and automated CI/CD (GitHub Actions + CodePipeline) for faster, safer releases. SKILLS

Languages: Python, TypeScript/JavaScript, SQL, Java,Golang Backend & Data: Node.js/Express, Django/FastAPI, PostgreSQL/MySQL, MongoDB, Redis Infra & ML: AWS, Docker, Kubernetes, CI/CD, Linux/Unix, Git, PyTorch/TensorFlow, pandas/NumPy

Contact this candidate