Senior Data & AI Platform Engineer

The CoinDCX Journey: Building the Future of Finance At CoinDCX, our mission is clear - to make crypto and blockchain accessible to every Indian and enable them to participate in the future of finance. As India's first crypto unicorn valued at 2.45B, we are reshaping the financial ecosystem by building safe, transparent, and scalable products that power adoption at scale. We believe that change starts together. It begins with bold ideas, relentless execution and people who want to build what's next. If you're driven by purpose and thrive in environments where your work defines the next chapter of an industry, you'll feel right at home here. About the Role Operating a premier crypto exchange means moving at the absolute speed of the market. Crypto never sleeps, risk patterns evolve continuously, and malicious actors iterate by the minute. At CoinDCX, our data engineering foundation is already highly mature processing billions of events daily via Databricks and Kafka. We aren't looking for someone to build basic data pipelines. We are looking for an exceptional engineer to construct the CoinDCX AI Value Platform. This horizontal infrastructure layer will transform our massive data footprint into automated, intelligent action. You will build the frameworks, model registries, and context stores that allow both classic machine learning models and state-of-the-art Agentic AI systems to execute critical workflows safely spanning real-time account takeover (ATO) containment, algorithmic crypto withdrawal risk-tiering, referral abuse detection, and AI-assisted wealth intelligence. You ll Excel in This Role If You Experience: 4+ years of intensive platform or data engineering experience. We are actively targeting seasoned SDE-2 or early SDE-3 candidates with elite programming fundamentals, massive learning velocity, and zero fear of shifting paradigms. The Core Stack: Expert-level mastery of Python, PySpark, and Spark SQL optimization. You should intimately understand how distributed memory management works and how to manipulate massive datasets efficiently. Streaming & Orchestration: Direct experience with Kafka/MSK and Apache Airflow (or Databricks Workflows) for complex, high-dependency system workflows. AI, MLOps & LLM Toolkit: Practical implementation experience with MLflow for production model lifecycles. Strong conceptual or practical exposure to Vector Architectures and LLM coordination abstractions (LangChain, LangGraph, or LlamaIndex). FinTech/Crypto Context (Massive Plus): Prior exposure to high-integrity transactional spaces such as order matching engines, double-entry ledgers, blockchain nodes, risk compliance systems, or real-time payment gateways is heavily favored. You ll Know You're Winning When MLflow Production Setup: Fully standardize and operationalize MLflow pipelines across the team, bringing our first set of live account takeover (ATO) and referral abuse detection models under structured lifecycle management. Entity 360 Rollout: Successfully ship the production data layers for User 360 and Wallet 360, cleanly feeding real-time context to upstream decision engines. Agentic RAG Infrastructure: Deploy the automated data ingestion, vector indexation, and evaluation framework for our digital customer support or internal intelligence agent. Zero SLA Deviations: Ensure all new AI Value Platform integrations dock cleanly into our billion-event stream without introducing data lag or compromising the stability of our transactional core. Engineer the CoinDCX Entity 360 & Semantic Layers Crypto Entity Architecture: Architect and optimize our Entity 360 Platform specifically unifying disparate data streams into high-performance, real-time context stores including User 360, Wallet 360 (On-chain/Off-chain balance states), and Token 360. Core Business Semantics: Build and govern a centralized Semantic and Metrics Layer to guarantee that data models, internal engines, and AI agents reference identical, deterministic definitions for core crypto metrics (e.g., active trader, malicious wallet cluster, referral loop, and crypto deposit/withdrawal (CDW) eligibility). Standardize Exchange-Scale MLOps & Lifecycle Tracking MLflow Platform Ownership: Own the deployment and standardization of MLflow (Model Registry, Tracking, Recipes) across CoinDCX to catalog, version, and deploy predictive models safely into our 24/7 production environment. Continuous AI Observability: Set up automated evaluation pipelines and tracing frameworks via MLflow LLM Tracking to capture live inputs/outputs, monitor data and feature drift, and benchmark model accuracy against real-world crypto market fluctuations. Build Agentic AI & Advanced LLM Infrastructure Agentic Orchestration: Design and scale the data-routing backends required for Multi-Agent Systems (using LangGraph, CrewAI, or similar frameworks) to automate intricate compliance and operational journeys such as auto-summarizing AML cases, evaluating token listing/delisting intelligence, and executing smart customer support agent routing. Production Context Engineering: Build low-latency Retrieval-Augmented Generation (RAG) data systems. Optimize data chunking strategies, embed generation, vector database indexing (via Databricks Vector Search), and semantic caching to eliminate hallucination vectors within our customer-facing applications. Leverage & Fuel Core Feature Stores Feature & Signal Stores: Build and maintain low-latency Feature Stores that pull directly from our live Databricks (PySpark, Spark SQL, Delta Lake) environments to serve unified real-time signals to downstream transaction-monitoring and threat-detection models. Streaming Platform Alignment: Interface seamlessly with our active Kafka/MSK, Auto Loader, and Change Data Capture (CDC) architectures to ensure your downstream AI applications scale effortlessly without impacting existing core ledger or reporting SLAs. Web3 Governance & Guardrails Enforce institutional-grade security guardrails directly into the platform layout: implement automated PII tokenization, wallet-masking policies, and rigorous access control via Unity Catalog. You ll Excel in This Role If You Experience: 4+ years of intensive platform or data engineering experience. We are actively targeting seasoned SDE-2 or early SDE-3 candidates with elite programming fundamentals, massive learning velocity, and zero fear of shifting paradigms. The Core Stack: Expert-level mastery of Python, PySpark, and Spark SQL optimization. You should intimately understand how distributed memory management works and how to manipulate massive datasets efficiently. Streaming & Orchestration: Direct experience with Kafka/MSK and Apache Airflow (or Databricks Workflows) for complex, high-dependency system workflows. AI, MLOps & LLM Toolkit: Practical implementation experience with MLflow for production model lifecycles. Strong conceptual or practical exposure to Vector Architectures and LLM coordination abstractions (LangChain, LangGraph, or LlamaIndex). FinTech/Crypto Context (Massive Plus): Prior exposure to high-integrity transactional spaces such as order matching engines, double-entry ledgers, blockchain nodes, risk compliance systems, or real-time payment gateways is heavily favored. You ll Know You're Winning When MLflow Production Setup: Fully standardize and operationalize MLflow pipelines across the team, bringing our first set of live account takeover (ATO) and referral abuse detection models under structured lifecycle management. Entity 360 Rollout: Successfully ship the production data layers for User 360 and Wallet 360, cleanly feeding real-time context to upstream decision engines. Agentic RAG Infrastructure: Deploy the automated data ingestion, vector indexation, and evaluation framework for our digital customer support or internal intelligence agent. Zero SLA Deviations: Ensure all new AI Value Platform integrations dock cleanly into our billion-event stream without introducing data lag or compromising the stability of our transactional core. Disclaimer: This job posting has been aggregated from external source. Role details, content, and availability are subject to change. Applicants are advised to confirm the latest information directly on the company website before applying.