Full-Stack Developer

We’re Zenith A boutique consulting firm based in Bangalore, New York and Dubai, with a global network of partner offices. We help startups build fractional finance teams — blending big-corporate sophistication with startup-speed execution. We’re seeking a Full-Stack Developer (AI/ML Focus) who can architect systems, ship production code, and apply engineering best practices. You’ll build features end-to-end across React/Next.js , AWS serverless services , and our multi-agent AI layer —while raising the bar on speed, reliability, security, and developer experience. Role Responsibilities: Product Engineering (Full-Stack) Build high-quality React/Next.js (App Router, Server Components) experiences with exceptional UX. Design and implement TypeScript APIs (rate limiting, idempotency, pagination, error taxonomy). Implement financial KPI/metrics engines and domain logic with rigorous correctness and tests. Integrate with QuickBooks/Zoho/banking APIs ; handle OAuth 2.0/token lifecycle securely. Delivery, Quality & DevEx (CI/CD) Establish CI/CD pipelines with trunk-based development, preview apps, and automated checks (type, unit, integration, e2e). Build a robust testing pyramid (Vitest/Jest, Playwright/Cypress , contract tests, load tests). Implement feature flags, migrations, and safe rollout/rollback strategies. Observability & Reliability (SRE-lite) Stand up metrics, logs, traces (e.g., OpenTelemetry), SLOs/SLIs. Add application-level health checks, circuit breakers, backoff/retry, and chaos/readiness probes. Handle capacity planning and cost/performance tuning across compute, storage, and LLM spend. Security & Compliance Drive secure-by-default practices (least privilege IAM, secrets management, encryption in transit/at rest). Implement data partitioning/tenant isolation; audit logging and anomaly detection. Enforce data privacy (PII minimization/retention), secure OAuth flows, and permissions/entitlements model. Data, AI & MLOps Productionize RAG pipelines (chunking, embeddings, vector stores) and prompt/tooling strategies. Build conversation memory and evaluation harnesses (hallucination/latency/quality metrics). Implement model routing, guardrails, and fallbacks; track token usage and cost. Establish offline/online evaluation loops and A/B testing for prompts/policies. Applied Machine Learning for Financial Analysis Build and productionize time-series models (cash-in/cash-out, revenue, burn, runway) using walk-forward validation, backtesting, and error budgets. Design anomaly & risk detectors (sudden expense spikes, delinquent AR, vendor anomalies) with streaming/batch inference and configurable thresholds. Ship tabular ML for propensity/risk scoring (payment default likelihood, churn risk, late-invoice probability). Develop document & entity ML : invoice/receipt parsing (OCR + layout/NLP), vendor/entity resolution, chart-of-accounts auto-classification. Platform & Architecture (Tech Leadership) Design scalable, secure architectures for multi-tenant B2B SaaS and real-time financial data flows. Contribute to decisions on patterns, frameworks, and data models; create lightweight ADRs and guardrails. Collaborate closely with product/design and customer teams; ship outcomes, not just features. Contribute to code reviews and pair sessions; uplevel code quality and documentation. What we’re looking for: 4–5 years building and scaling modern SaaS (0→10K+ users). (This role isn’t open to freshers or early juniors.) Bachelor’s degree or higher  in Computer Science, Engineering, or a related field. Expert in Next.js (App Router, Server Components, API Routes) and advanced TypeScript/React patterns . Strong Python for backend & ML (FastAPI, asyncio) ; clean, typed, test-driven code. Production AWS (Lambda, API Gateway/ALB, DynamoDB, S3, CloudWatch; Cognito, EC2 preferred ). Real-time data (WebSocket/SSE ), distributed caching (Redis/Upstash ), OAuth 2.0. Hands-on LLM integration (OpenAI/Anthropic/etc.) in production. Applied ML (Finance) : time-series forecasting and risk/anomaly scoring with rigorous walk-forward/backtesting; plus doc/entity ML (e.g., invoice OCR or COA auto-classification). MLOps & Data Platform : pipelines with feature store, model registry, versioned data, CI/CD, and drift monitoring across batch/stream (e.g., MLflow or Kinesis). Fintech/accounting domain; QuickBooks/Xero/Zoho integrations. Agent frameworks; vector DBs; PDF/reporting pipelines. Data visualization (Chart.js/Recharts/ECharts/D3 ). Thrive in a high pressure and fast paced startup environment — fast-moving, collaborative, and comfortable with autonomy. Role Details: Start date : Immediate | Full-time Location: Hybrid (Bangalore)  Exposure: Shape the voice and content strategy for both a premier consulting brand and an innovative AI product from the ground up. Impact: Own the entire content experience for our digital ecosystem from day one, with your work directly driving our market position and customer acquisition. Compensation: Competitive salary with performance-based growth