Raghavendra
Singh Chauhan
Backend Engineering · Multi-Agent Workflows · Security Infrastructure
I build resilient, distributed platforms and agentic workflows that operate deterministically at scale.
Most recently, I architected a multi-framework compliance automation and EDR platform at an early-stage cybersecurity startup, leveraging DAG-based mappings and autonomous multi-agent pipelines to scale to thousands of tenants. Before that, I spent 2.5 years at Deutsche Bank engineering high-throughput microservices and optimizing heavily-loaded JVM clusters.
I am drawn to systems that demand deep reasoning—whether that means scaling retrieval-augmented generation (RAG) over vast vector topologies, orchestrating asynchronous training pipelines, or taking a deliberate sabbatical to study pure mathematics for India's civil services exam (ranking in the top 1% of 800,000+ candidates).
SDE-2
- Engineered an evidence-first compliance automation engine by building reusable SQL evidence collectors and a DAG-based control model, reducing duplicate evidence collection across multiple compliance frameworks.
- Designed and implemented backend services for an Endpoint Detection & Response (EDR) platform, enabling secure telemetry ingestion with resilient event processing, automated mTLS certificate management, and fault recovery through DLQs.
- Built AI-assisted engineering workflows by integrating Cloud MCP servers with Jira and Confluence, automating documentation, ticket synchronization, and development lifecycle tasks to improve engineering productivity.
Deliberate Sabbatical
Chose to compete for India's most selective examination with Mathematics as the optional subject. Cleared Civil Services Prelims — Top 1% of 800,000+ candidates. Deep study of Linear Algebra, Analysis, and Numerical Methods.
Software Engineer (Associate)
- Fast-tracked to Associate (SDE-2) for consistently exceeding expectations, independently driving modernization initiatives, and mentoring junior engineers.
- Built and scaled Spring Boot data-ingestion microservices processing 50M+ records/day using MQ and Spark with 80%+ automated test coverage and mTLS-secured communication.
- Co-led modernization of the Gravity platform by decomposing a legacy monolith into an event-driven ingestion pipeline, delivering approximately $500K annual infrastructure savings.
- Owned JVM performance tuning for a 24-node Hazelcast cluster, optimizing G1GC configuration and eliminating GC-induced latency during high-volume cVA risk calculations.
- Optimized Drools rule evaluation using HashMap-based indexing and incremental loading, reducing rule execution latency by nearly 40%.
- Led the Hazelcast 4.x & Apache Camel 3.x upgrade across a tightly coupled monolith, preventing an estimated 8-week regulatory compliance delay.
- Resolved production incidents involving 160M+ financial records, identifying data corruption root causes and restoring business-critical processing under strict regulatory timelines.
- Built an agentic lead intelligence platform automating discovery, enrichment, and qualification, processing 10k+ businesses/day using autonomous LangGraph workflows.
- Integrated asynchronous data pipelines and external business APIs, reducing end-to-end lead enrichment latency by 70%.
- Engineered a distributed ingestion pipeline indexing 100k+ documents using Virtual Threads, Kafka fan-out, and LangChain4j embeddings.
- Implemented hybrid semantic retrieval with pgvector (HNSW) and Redis, achieving sub-100 ms contextual search latency for complex RAG queries.
- Built a distributed AI training platform supporting 1,000+ concurrent training jobs through asynchronous task orchestration.
- Developed diffusion-based synthetic data generation for datasets containing 10M+ rows using memory-efficient streaming pipelines.
IIEST Shibpur
B.Tech, Electronics & Telecommunications
West Bengal, India · 2017–2021