Headline

I build scalable backend systems | Software Engineer specializing in cloud-native services • Expertise in Node.js, Python & AWS • Experience with event-driven architectures

About

I’ve always been fascinated by how technology can solve real-world problems. My journey into the world of backend systems started during my time at university where I built a strong foundation in data science and programming. However, it was through my internships that I truly discovered my passion for crafting scalable systems that can handle complex demands. Each experience has shaped me into the software engineer I am today. At Spectatr.ai, I had the incredible opportunity to architect an end-to-end live stream ingestion pipeline using AWS MediaLive that processes over 50+ concurrent streams. This project not only reduced stream-to-playback latency to under 8 seconds but also enhanced user experience significantly. It was thrilling to see how our work directly impacted users enjoying seamless streaming. Before this role, I interned at Samsung R&D Institute India where I tackled critical compatibility issues across 15+ device models. By debugging effectively, I managed to reduce the Android app crash rate from 8.2% to 6.1%, a 25% improvement! This experience taught me the importance of attention to detail and how even small changes can lead to substantial improvements. My earlier stint as a Software Developer Intern at Voorent allowed me to dive into feature development using technologies like Springboot Java and React.js while ensuring efficient database maintenance. It was rewarding to see how my contributions streamlined processes within the team. Outside of work, I enjoy sharing my knowledge through technical writing. During my time as a Technical Content Writer for The Crazy Programmer, I produced 50 articles on various programming languages which helped demystify complex concepts for readers. This blend of technical expertise and communication fuels my desire to continue learning and growing in this field. If you’re interested in discussing backend systems or cloud-native services, or if you just want to share insights about tech, feel free to reach out via email! Skills: - Node.js - Typescript - Python - AWS - Event-driven architectures - Java - System Design - Data Structures & Algorithms

Location

Bengaluru

AI stack

anthropic apis

claude code

claude

cursor

gemini

langchain

Work experience

Software Development Engineer | AWS Cloud Services, Kafka Streaming, Event-Driven Architecture at SPECTATR.AI, Jul 2024 - Jul 2026 · 2 yrs 1 mo, Gurugram, •Architected end-to-end live stream ingestion pipeline (AWS MediaLive → S3 → Lambda → ECS) processing 50+ concurrent streams, reducing stream-to-playback latency to under 8 seconds and eliminating 100% of duplicate processing failures (previously 15–20 daily) via Redis distributed locking. •Reduced live commentary latency from 3.5s to 500ms (85% improvement) using AWS Transcribe and SQS-based workload isolation, while cutting transcription costs from ₹1.7L to ₹1K monthly (99% reduction) via on-demand Fargate scaling. •Achieved 99.9% message delivery reliability (zero data loss across 2M+ daily events) by architecting 4-topic Kafka streaming pipeline with auto-scaling ECS consumers per partition, processing real-time video analytics for 10K+ concurrent viewers. •Reduced content authentication latency by 60% (from 450ms to 180ms per request) by implementing CloudFront signedcookie system, eliminating per-asset URL signing overhead for 5K+ daily video requests while maintaining enterprise-grade security. •Reduced malicious and out-of-scope LLM requests by 17% (filtering 8,500+ monthly queries) through guardrails agent implementation, saving ~₹45K monthly in inference costs and preventing hallucinated responses from reaching users. •Reduced LLM hallucination rate from 32% to 6% (81% improvement) by implementing LangSmith evaluation framework with hallucination-detection judges and feedback-driven prompt refinement, improving production response accuracy for 15K+ daily queries. •Built multi-platform social publishing system handling 100+ daily posts across 5 channels — 98% delivery success rate with 85% failed-attempt recovery via SQS exponential backoff, eliminating manual intervention for 500+ weekly publishes. •Resolved critical folder deletion race condition (40+ daily data corruption incidents) using BullMQ batch processing with Redis atomic locking, improving storage calculation accuracy from 26% to 99% and eliminating manual engineering cleanup.

Software Engineer Intern | Android Development, UI Automation, and Compatibility Debugging at Samsung R&D Institute India, Jan 2024 - Jun 2024 · 6 mos, Noida, Key accomplishments include: * Reduced Android app crash rate from 8.2% to 6.1% through debugging and resolving over 30 critical compatibility issues across 15+ device models, enhancing user retention for over 500K active users. * Developed an automated UI testing framework utilizing Espresso and UI Automator with over 200 test cases, leading to a 40% decrease in regression defects from 15 to just 9 per release cycle. * Standardized testing frameworks across four engineering teams comprising more than 30 developers, boosting test coverage from 65% to 85% while cutting down production hotfixes by 35%. * Conducted quarterly review sessions that documented best practices for quality assurance processes, fostering a culture of continuous improvement within the engineering teams. Skills: Android Development, Automated Testing Frameworks, Compatibility Debugging, User Retention Strategies.

Software Developer Intern | Springboot Java, React.js & AWS Cloud Services at Voorent, Jul 2022 - Jan 2024 · 1 yr 7 mos, Gurugram, Key accomplishments include: * Developed multiple new features using Springboot Java and React.js that enhanced user experience by 30%, leading to increased user engagement and positive feedback from stakeholders. * Resolved over 50 critical bugs in the application while maintaining efficient database operations with Postgres, resulting in a 25% decrease in error reports and improved system stability. * Streamlined deployment processes utilizing Docker and AWS services which reduced deployment time by 40%, allowing for faster updates and feature releases to the production environment. * Created comprehensive invoice generation systems that automated financial operations and improved accuracy by 95%, contributing to seamless operational workflows for the finance department. Skills: Springboot Java, React.js, Postgres, Docker, AWS

Technical Content Writer at The Crazy Programmer, May 2021 - Jul 2021 · 3 mos, Udaipur, Rajasthan, India, • Produced 50 informative articles on diverse programming languages and tech-focused subjects as a Technical Content Writer. • Demonstrated expertise in creating engaging content encompassing technical concepts for readers. • Enhanced readers' understanding of programming languages through clear and concise writing.

Education

B.Sc, Data Science and Programming, Indian Institute of Technology, Madras

Bachelor of Engineering - BE, Computer Science, CHANDIGARH UNIVERSITY

Public links

github: https://github.com/agarNit

leetcode: https://leetcode.com/agar_Nit

youtube.com: https://youtube.com/@NitishBuildsLife

codeforces: https://codeforces.com/profile/agarwalnitish121