Ankur Rathore

Role: Independent Systems Engineer & Open Source Developer
Tech Stack: Rust, eBPF (Aya/XDP), Linux Kernel ABI, Data-Oriented Design, Memory Management

Outside of my enterprise architecture roles, I am deeply passionate about solving extreme performance bottlenecks in cloud-native and AI-inference environments. My open-source research focuses on bypassing high-level overhead to deliver zero-cost, kernel-level networking and observability tools natively in Rust.

Key Open-Source Projects:

• eBPF Kernel Firewall (xdp-ai-guard): Built a programmable firewall running directly inside the Linux Kernel using Rust and eBPF (XDP) to protect AI inference servers from volumetric DDoS attacks. Achieved line-rate packet filtering and stateful rate-limiting by executing logic natively in the network driver, preventing sk_buff allocation memory exhaustion.
• Zero-Dependency Linux Observability (vmstat-rs): Engineered a high-frequency Linux monitoring agent in Rust. Bypassed standard libraries to manually parse the /proc ABI, achieving sub-millisecond precision with <1% CPU overhead. Implemented real-time OOM Kill detection, Context Switch alerting, and granular memory forensics (Major vs. Minor Page Faults) to detect disk I/O bottlenecks stalling GPU pipelines.
• High-Performance Memory Optimization (sharded-timing-wheel): Engineered a custom Slab Allocator (Arena) utilizing intrusive linked lists in Rust to eliminate heap fragmentation and reduce pointer-chasing overhead. Implemented a Hierarchical Timing Wheel to replace standard O(log⁡N)
priority queues with O(1)constant-time scheduling, targeting C10M-scale network driver use cases via CPU cache pre-fetching.

Role: Technical Lead & Systems Architect
Tech Stack: Python, Apache NiFi, Kafka, Kubernetes, Rancher, CrateDB, FIWARE, PySNMP, C#

As a Technical Lead at NEC, I architect high-throughput distributed systems and data pipelines for Enterprise Smart City and Telecom IoT platforms, bridging raw hardware telemetry with scalable cloud backends.

Key Architectural Achievements:

• IoT Data Orchestration: Architected a resilient, event-driven ingestion engine using Apache NiFi and Kafka to buffer unpredictable telemetry spikes (mitigating "thundering herd" issues) from thousands of heterogeneous sensors.
• Microservices & Normalization: Built asynchronous Python microservices to consume, validate, and normalize proprietary hardware payloads into strict NGSI standards, routing data to distributed time-series databases (CrateDB).
• Platform Automation: Engineered a Python-based automation framework that replaced manual legacy device provisioning, reducing IoT device onboarding time by 95%.
• Telecom Network Automation: Led the technical implementation of Digital Longitudinal Monitoring to track Signal-to-Noise Ratios (SNR) for NEC 800G optical transponders in real-time. Designed a Python/PySNMP backend to automatically discover and map physical network nodes (L2/L3 Switches/Routers).
• Infrastructure & SRE: Containerized and managed mission-critical deployments across distributed Kubernetes and Rancher clusters. Provided Tier-3 production SRE support, resolving complex orchestration and networking bottlenecks.
• Low-Latency Biometrics: Developed a custom C# SignalR client to integrate real-time facial recognition streams into the Python/NiFi ecosystem for instantaneous security processing.

Role: Senior Software Engineer & Team Lead
Tech Stack: Python (Django), Go, React, AWS (Lambda, Step Functions), Kafka, SDN, Cisco APIs

At Intuitor, I engineered high-performance cloud automation platforms and enterprise SaaS portals, bridging the gap between legacy telecom infrastructure and modern AWS serverless architectures. I operated as both a core backend contributor and a technical team lead.

Key Architectural Achievements:

• Serverless Hyper-Automation: Architected core modules for a SaaS-based AWS security platform. Leveraged AWS Step Functions and Lambda to orchestrate complex state machines and remediation workflows, successfully automating 90%+ of manual security tasks for enterprise clients.
• High-Performance Go & Kafka Eventing: Engineered a high-speed, concurrent CLI tool in Go for event-driven resource management. Integrated Kafka to reliably process and buffer real-time security alerts under heavy load without dropping messages.
• Software Defined Networking (SDN) Leadership: Led a team of 3 engineers to build a unified cloud resource provisioning engine. Abstracted proprietary hardware protocols across Megaport, Equinix, and Nokia routers into clean REST APIs, drastically reducing resource turnaround time by 60%.
• Full-Stack Enterprise Portals: Architected and delivered end-to-end web applications, including a Network Inventory Management Portal and a Cisco IP Telephony Portal. Built robust backend APIs with Python/Django and dynamic frontends using React.
• Real-Time Systems Monitoring: Developed a custom real-time telemetry and monitoring system for Cisco/Polycom video conferencing, implementing automated call quality analysis and troubleshooting features.