What You’ll Do
- Own features end-to-end: from API & data models to responsive UIs and observability.
- Ship agentic AI: design tool-using agents (planning, tool orchestration, retries/guardrails), RAG pipelines, and evaluations to improve task success.
- Build scalable, secure platforms: low-latency services, event-driven workflows, and defense-in-depth for PCI/PII.
- Develop AI/ML capabilities: model integration, prompt/flow engineering, embeddings, vector search, online inference, and A/B evaluations.
- Optimize performance: reduce p99 latencies, cost, and flakiness across services and frontends.
- Collaborate cross-functionally: product, design, DS/ML, risk/fraud, and compliance.
- Mentor and uplift: review PRs, improve patterns, document best practices.
Tech You May Use Here
- Frontend: React/Next.js, TypeScript, state mgmt (Redux/Zustand), Tailwind, component libraries; testing (Jest/RTL), Webpack/Vite.
- Backend: Java (Spring Boot), Python (FastAPI), Node.js/TypeScript (NestJS); REST/GraphQL/gRPC; WebSockets.
- Data & Infra: Postgres/MySQL, Redis, Kafka/PubSub, object storage; CI/CD (GitHub Actions/Azure DevOps), containers; AWS/Azure/GCP.
- AI/ML & Agentic: OpenAI/Anthropic/Gemini SDKs, function/tool calling, LangChain/Autogen/CrewAI, vector DBs (pgvector/FAISS/Pinecone), evals; Copilot ecosystem / Copilot Studio, Power Apps/Power Automate; MLOps (Dockerized inference, model/versioning, feature stores).(Must to have)
- Security/Compliance: secrets management (Key Vault/SM/Secrets Manager), authN/Z (OIDC/OAuth2), audit, rate limiting, PCI considerations.
Responsibilities (Split by Strength)
- If you’re backend-leaning (primary):
- Design and evolve domain services, APIs, and data models for cards/payments.
- Build resilient, observable services (metrics, tracing, alerts); capacity planning and on-call participation.
- Integrate AI services and agents into backend workflows (tool routers, retrieval, safety/guardrails).
- If you’re frontend-leaning (primary):
- Build accessible, performant React/Next.js apps with clean state and design systems.
- Implement real-time UX for risk/fraud dashboards, ops consoles, and customer features.
- Integrate agentic-AI UX (multi-step plans, tool traces, explanations, feedback loops) with secure server boundaries.
Minimum Qualifications
- 3-5 years professional experience (or equivalent) building production software.
- Excellent proficiency in either frontend or backend, and working ability across the other side.
- Solid CS fundamentals (DSA, concurrency, network & HTTP, caching, DB design).
- Proven experience shipping AI/ML or agentic-AI projects (e.g., RAG/embeddings, tool-using agents, production model integration, or Copilot Studio solutions).
- Experience with cloud (AWS/Azure/GCP), containers, CI/CD, and observability.
- Strong problem-solving, ownership, and clear written/verbal communication.
Nice to Have
- Payments/banking domain exposure (fraud, risk, ledgering, reconciliation) or high-throughput systems.
- Knowledge of GNNs, NLP, or AI-powered fraud signals.
- MLOps (model deployment, batch/online features, canarying, eval harnesses).
- Experience with Microsoft ecosystem (Teams apps, Graph, Copilot Studio/Power Platform).
- Performance tuning (bundles/pain points on FE; JVM/async IO on BE); accessibility.
What Success Looks Like (First 90 Days)
- Ship a meaningful feature end-to-end (API → UI) with observability and docs.
- Integrate or improve an agentic flow (tool orchestration + evaluation) that measurably improves a KPI (e.g., time-to-resolution, approval rate, ops effort).
- Reduce a critical p99 or error rate by a concrete target through profiling and design fixes.
- Raise the bar via reviews, refactors, and reusable components/patterns.
Top Skills
What We Do
Founded in 2015, Zeta is a provider of next-gen credit card processing platform. Zeta’s cloud-native and fully API-enabled stack offers a comprehensive range of capabilities, including processing, issuing, lending, core banking, fraud detection, and loyalty programs. With a strong focus on technology, Zeta has over 1700+ employees and contractors, with more than 70% dedicated to technology roles. Operating across the US, UK, Middle East, and Asia, Zeta has served a global customer base of 35+ clients who have issued over 15 million cards on Zeta's platform to date. Backed by prominent investors such as Softbank Vision Fund 2 and Mastercard, Zeta has raised $280 million, at a valuation of $1.5 billion.