Sunney S.
Results-driven AI, Data Engineering, and Backend Software Engineer with 5+ years of experience designing, developing, and scaling distributed systems, cloud-native microservices, and production-grade data pipelines. Strong expertise in AWS, GCP, LLM-powered applications, Generative AI, multi-agent architectures, and Retrieval-Augmented Generation (RAG). Experienced in building high-performance backend systems and processing large-scale data to deliver reliable, scalable, and intelligent solutions. Proven ability to transform complex data and business requirements into end-to-end solutions, from data ingestion and processing to AI-driven insights and decision-making.
- Architect and develop production-grade multi-agent LLM systems, integrating third-party APIs to deliver intelligent, scalable, and user-facing AI solutions.
- Design and optimize Retrieval-Augmented Generation (RAG) pipelines for accurate information retrieval from complex financial and lending documents, improving response quality and retrieval precision.
- Lead end-to-end system architecture, technical design, and development workflows across multiple product features, establishing engineering best practices for AI-driven applications.
- Research and implement advancements in LLM agent frameworks, model alignment, and AI security to enhance the reliability, safety, and robustness of production systems.
- Drive continuous improvements in system performance, maintainability, scalability, and security through iterative refactoring, optimization, and code-quality initiatives.
- Built and maintained scalable ETL pipelines using Apache Airflow, ensuring reliable data ingestion, transformation, and delivery for downstream analytics and data products.
- Designed, developed, and deployed scalable REST APIs and microservices using Django, Flask, Django REST Framework, and FastAPI, enabling seamless integration with third-party services and proxy providers.
- Independently developed an end-to-end profile-based Google search result scraping platform, covering data acquisition, complex data-cleaning workflows, and loading processed datasets into Google BigQuery.
- Engineered and optimized a high-performance HTML parsing system using Celery for asynchronous processing, improving throughput across multi-million-page datasets.
- Applied LLMs and advanced prompt-engineering techniques to HTML parsing and sentiment-analysis workflows, reducing manual review effort and improving content-classification accuracy.
- Leveraged a broad range of AWS services, including EC2, S3, Redshift, SQS, ECS, Athena, Bedrock, Lambda, EventBridge, and Secrets Manager, to build scalable cloud infrastructure, serverless workflows, and automated processing pipelines.
- Led the migration of AMD-based cloud infrastructure to ARM architectures, delivering significant cloud-cost savings at scale while maintaining system performance and reliability.
- Worked extensively with PostgreSQL, Amazon Redshift, Google BigQuery, MongoDB, and Redis to support diverse data-storage, processing, and modeling requirements across multiple product domains.
- Developed and implemented RESTful APIs using FastAPI, with SQLAlchemy ORM and Alembic migrations, for a Dutch trucking and logistics platform serving Redistri BV.
- Automated business invoice generation using ReportLab, eliminating manual document creation and streamlining the billing process.
- Integrated AWS Simple Email Service (SES) for transactional and templated email delivery, improving automated communication and operational workflows.
- Developed hands-on expertise in Python, Pandas, and PostgreSQL while contributing to multiple ETL projects involving data ingestion, transformation, pipelining, and warehousing.
- Contributed to data acquisition and processing workflows, supporting downstream analytics, reporting, and data-driven applications.