Swathi Sadasivan
Data Scientist & AI Engineer with 5+ years of experience building secure back-end systems, high-throughput data pipelines, and production-grade AI applications. Specialized in deploying low-latency RAG architectures, PySpark ETL, multi-agent systems, and cloud infrastructure across AWS & Azure.
Swathi Sadasivan
MSc Data Science & Analytics
Ex-DataStax (IBM) | Ex-IBS Software
Technical Expertise
Specialized engineering competencies across Data Science, GenAI, and Cloud Infrastructure
Generative AI & LLMs
- RAG ArchitecturesProduction
- Vector DatabasesAstra DB
- LLM OrchestrationLangChain
- AI Multi-AgentsFastAPI
- Prompt EngineeringOpenAI
Data Engineering
- Distributed ETLPySpark
- StreamingApache Kafka
- LakehouseDatabricks
- IntegrationApache Camel
- Data GovernancePurview/Unity
Cloud & DevOps
- AWS CloudLambda, S3, DynamoDB
- Microsoft AzureEnterprise
- CI/CD AutomationGitHub Actions
- Backend StackSpring Boot, Java
- MethodologyTDD, Maven/Gradle
Data Science & ML
- Ensemble MLXGBoost, Random Forest
- Deep LearningCNN, TensorFlow
- Imbalanced DataSMOTE
- Model TuningCalibration & SciKit
- AnalyticsPower BI, SQL
Featured Engineering Projects
Production-grade AI systems, high-throughput pipelines, and predictive ML models
Enterprise Knowledge Assistant (RAG System)
Built a production-ready, low-latency Retrieval-Augmented Generation assistant capable of delivering sub-second semantic retrieval and response generation over a repository of 50,000+ enterprise documents.
- Engineered efficient document chunking, embedding generation & context window optimization.
- Utilized Astra DB Vector Search for vector indexing with FastAPI endpoint layer.
Real-Time Streaming & Distributed ETL Pipeline
Constructed high-throughput, event-driven streaming data pipelines using Apache Kafka to ingest real-time transactional data streams into downstream analytics warehouses.
- Developed automated ETL orchestration in PySpark for schema validation & data cleansing.
- Integrated microservices via Spring Boot & Databricks analytical aggregations.
Predictive Modeling & Anomaly Detection
Developed an end-to-end machine learning pipeline utilizing Kaggle's Blood Anomaly Detection dataset to reliably detect anomalous classes with severe distribution skew.
- Addressed extreme class imbalance with SMOTE to generate balanced training profiles.
- Fine-tuned Random Forest & XGBoost classifiers with probability calibration.
Deep Learning Image Classification
Designed and implemented custom Convolutional Neural Network (CNN) architectures to perform computer vision classification on MNIST and Fashion-MNIST datasets.
- Conducted experimentation comparing Adam vs. SGD optimization convergence rates.
- Achieved 85%+ accuracy using strategic batch sizing & dynamic epoch scheduling.
Professional Experience
5+ years leading data engineering, GenAI architectures, and cloud solutions
Data & Generative AI Engineer
Independent AI & Data Engineering Consultant
- Architect and develop production-grade ETL and AI-driven data pipelines using Python, PySpark, and SQL to process enterprise datasets across AWS and Azure.
- Design and deploy robust RAG applications powered by LLMs, intelligent AI agents, and production vector databases.
- Implement observability frameworks including monitoring, tracing, and logging to track retrieval accuracy and eliminate model hallucinations.
- Automate infrastructure provisioning via GitHub Actions using cloud-native IAM security and governance configs.
Cloud Solutions Engineer - AI & Data Platforms
DataStax (Acquired by IBM)
London, UK
- Architected enterprise Generative AI and data platform solutions utilizing Astra DB Vector Search, Python, Java, AWS, and Azure for high-accuracy RAG workflows.
- Built technical proofs-of-concept (PoCs) and end-to-end data pipelines incorporating LLMs, LangChain, and multi-agent systems.
- Led technical enablement workshops and architecture reviews around data governance, security, and metadata tracking with Unity Catalog and Microsoft Purview.
Senior Software Engineer
IBS Software
India
- Engineered high-performance data microservices utilizing Spring Boot, Python, and AWS architecture (Lambda, S3, DynamoDB), reducing query latencies by 35%.
- Designed automated CI/CD deployment pipelines using GitHub Actions and TDD, boosting test coverage by 45%.
- Extracted business insights from relational infrastructure using optimized SQL queries and Power BI dashboards.
Software Developer - Backend & Systems
UST Global
India
- Built real-time event streaming pipelines for healthcare analytics utilizing Java, Spring Boot, Apache Kafka, Apache Camel, and SQL.
- Optimized database queries and indexing strategies, dropping latencies by 50% while scaling capacity to safely handle 500,000+ daily transactions.
Live GitHub Open Source Integration
Fetch and explore live public repositories directly via GitHub REST API
Education
MSc in Data Science and Analytics
Royal Holloway University of London, UK
Core Focus: Machine Learning, Deep Learning, Data Mining, Statistical Analysis, Data Engineering, Big Data Analytics.
B.Tech in Computer Science and Engineering
University of Kerala, India
Certifications & Awards
Certifications
- Google AI Essentials
- HackerRank SQL (Advanced)
- Data Engineering Foundations
- Enterprise Data Governance & Data Science Bootcamp
- Automation Pro II (Workato)
Key Awards
Let's Connect & Collaborate
Looking for a Senior Data & AI Engineer to build low-latency GenAI services, event-driven pipelines, or enterprise vector search infrastructure? Let's discuss your engineering goals.