Staff Backend Engineer
PHANEENDRA
Kovai Systems India Private Limited
Bengaluru
14
Years of experience
Summary
Strategic, hands-on Staff Backend Engineer & Technical Architect with 14+ years of professional experience designing production-grade, high-throughput Microservices Architectures and highly resilient distributed systems across enterprise product, banking, and AI orchestration platforms. Proven mastery engineering stateful task execution layers, asynchronous job queues, and fault-tolerant multi-agent topologies using Java, Spring Boot, Kafka, Temporal, Qdrant, and PostgreSQL. Expert in balancing core system design constraints—such as high availability, strict data consistency, and low-latency p99 performance profiles—with cutting-edge LLMOps and FinOps infrastructure, managing the complete technical lifecycle from initial system architecture to global production scale.
Work History
Technical Lead Architect / SDE-3 (Architect Track)
1 Year 11 Months- Kovai Systems acquired Floik, the product built at Ping Technologies — same product and team, under a new parent company.
- Projects: Floik, Document360 Integration
- Led an agile team of 5 engineers to architect a secure, distributed Multi-Agent Topology; built highly resilient, horizontally scalable microservices using Java, Spring Boot, Kafka, and Temporal (Infrastructure Orchestrator) to handle stateful task planning, Human-In-The-Loop (HITL) Verification gates, and deterministic recovery of production workflows.
- Engineered advanced enterprise RAG & Ingestion Strategies featuring Recursive Character Chunking and syntax-aware code parsing; implemented a high-performance Retrieval & Reranking Architecture combining Dense Vector Search (Qdrant) and Sparse Keyword Search (BM25) aligned via Reciprocal Rank Fusion (RRF), validating pipeline performance using the Ragas framework (Context Recall & Faithfulness metrics).
- Architected a real-time AI Data Pipeline Orchestration layer utilizing Change Data Capture (CDC) to auto-embed and sync high-volume operational databases into partitioned Qdrant collections with strict data isolation, structural metadata filtering, and optimized replication factors.
- Maintained p99 processing latencies under strict baselines by managing media processing pipelines processing 10K+ concurrent video tasks daily; leveraged fine-tuned database routing and bounded parallel execution loops to handle sudden infrastructure load shifts.
- Acted as cross-functional anchor, unblocking frontend architectural milestones during resource constraints by temporarily operating as a full-stack React engineer to deploy, test, and ship crucial application views.
Technical Lead Architect / SDE-3 (Architect Track)
2 Years 10 Months- Projects: Floik (storytelling platform generating visually rich interactive product demos, videos, and step-by-step guides)
- Contributed core platform engineering and high-availability architecture for Floik's high-visibility launch, maintaining fault-tolerant system health to capture the 2023 Golden Kitty Award and #1 Product of the Day/Week validation.
- Architected an intelligent LLMOps Model Router & Fallback proxy across multi-provider cloud layers (OpenAI/Anthropic/Azure); integrated a high-speed Semantic Caching Engine (RedisVL) to catch redundant prompts, achieving instantaneous failover boundaries and slashing API token bills.
- Engineered secure, high-throughput RESTful gateways embedded with custom FinOps Cost Allocation tables , multi-tenant token tracking, Context Window Optimization (dynamic history trimming), and token-bucket Rate Limiting to seamlessly govern shared enterprise API keys.
- Guaranteed a 99.99% service availability baseline by designing resilient event-driven microservices anchored on CAP theorem principles, active failure-isolation zones, and strict data idempotency patterns.
- Performed deep-dive Java performance tuning and JVM profiling; optimized concurrency libraries, multi-threaded worker pools, and garbage collection (GC) profiles to systematically eliminate memory leaks, CPU saturation, and transient latency bottlenecks.
- Established comprehensive testing infrastructure (JUnit, Mockito, and automated contract testing blocks) paired with automated DevOps execution (Kubernetes, AWS EKS, Jenkins, GitHub Actions, and SonarQube quality gates) to protect core builds against regression errors.
Software Developer - SDE-3
9 Years 7 Months- Projects: Razorthink aiOS, FormAI, Customer Churn, Next Best Product, Demand Forecasting, Adonme, Anounze, Guest Verification Portal, Incident Management System, OurUpgrade, SafetyStay
- Directed data validation, analysis, and preprocessing architecture for Customer Churn, Demand Forecasting, and Next Best Product initiatives in the banking sector.
- Designed a multi-module, large-scale data processing engine and AI platform supporting batch and real-time processing; constructed automated DAG infrastructure and data-ingestion task planners to optimize heavy asynchronous pipelines.
- Defined process improvements that reduced data-preprocessing cycle times.
- Performed Java performance tuning and JVM profiling, optimizing concurrency, multithreading, garbage collection (GC), and JVM configurations to identify and resolve CPU, memory, thread, and latency bottlenecks and improve system throughput.
- Enforced rigorous code quality via comprehensive JUnit, Mockito, and contract testing layers to maximize regression coverage.
Education
Bachelor of Technology (B.Tech) - Computer Science and Engineering (CSE)
Skills
Architecture & Core Design: System Design
Microservices Architecture
Event-Driven Architecture
High-Availability Distributed Systems
Fault-Tolerant Topology
Async Job Queues
Task Planners
DAG Engines
CAP Theorem Zoning Enterprise AI & Orchestration: Agentic Systems
Multi-Agent Topologies
State Management
Memory Engineering (Sliding Window
Episodic Memory)
Model Context Protocol (MCP)
LangGraph
LlamaIndex
LangChain
LLMOps & FinOps Infrastructure: Model Routers
Multi-Provider Fallback Proxies
Semantic Caching (RedisVL)
Context Window Optimization
Multi-Tenant Token Budgeting
Rate-Limiting Cost Allocation
RAG Evaluations (Ragas
Context Recall
Faithfulness)
Languages & Frameworks: Java (8/11/17/21)
Python
Spring Boot
FastAPI
React (TypeScript)
Core Frameworks & SDKs
Data
Messaging & Vector: Apache Kafka
ZMQ
Redis
Qdrant (Vector DB Architecture)
PostgreSQL (Relational Analytics)
MySQL
MongoDB
Elasticsearch
Change Data Capture (CDC Pipeline Orchestration)
Cloud
DevOps & Infrastructure: AWS (EKS
Lambda
S3
MSK
CloudFront
EventBridge
KMS
MediaConvert)
GCP
Docker
Kubernetes
CI/CD Automation (Jenkins
GitHub Actions)
SonarQube
JVM Concurrency Profiling
API design and management
Software architecture design
Stakeholder communication
Frontend and backend development
ARCHITECTURE HIGHLIGHTS
- Award-Winning Distributed Infrastructure: Designed the high-throughput, event-driven distributed infrastructure powering Floik's core engine, achieving #1 Product of the Day, #1 Product of the Week, and winning Product Hunt's 2023 Golden Kitty Award.
- High-Throughput Concurrency Scale: Scaled media processing and ingestion pipelines to handle 10K+ concurrent records/day, lowering p99 processing latency by 30% through precise chunk-based processing configurations, database routing, and bounded async concurrency loops.
- Deterministic Distributed State Machines: Deployed highly resilient, fault-tolerant execution pipelines utilizing LangGraph and Temporal, establishing strict data idempotency patterns and stateful orchestration layers that guarantee 100% state recovery under extreme traffic surges.
- Secure Enterprise API Gateways: Architected production-grade Model Context Protocol (MCP) servers alongside custom middleware interceptors to expose monolithic core APIs as strongly-typed, secure tools, enforcing runtime compliance via PII Detection, Prompt Injection Defence, and Human-In-The-Loop (HITL) Verification gates.
- High-Scalability Vector Indexing: Optimized distributed vector storage and advanced RAG search frameworks, increasing autonomous task accuracy by 25% using high-scalability Qdrant vector partitioning, metadata filtering, and PostgreSQL analytics lookup loops.
Timeline
Technical Lead Architect / SDE-3 (Architect Track)
Kovai Systems India Private Limited
11.2024 - Current
Technical Lead Architect / SDE-3 (Architect Track)
Ping Technologies
01.2022 - 11.2024
Software Developer - SDE-3
Razorthink (Technologies & Software)
06.2012 - 01.2022
Sri Sivani College of Engineering, JNTU-Kakinada
Bachelor of Technology (B.Tech) from Computer Science and Engineering (CSE)