funinkina.co.in · hello@funinkina.co.in · github.com/funinkina · linkedin.com/in/funinkina
Backend and systems engineer in Go, Python, and C/C++, with production experience across real-time LLM infrastructure, voice agent backends, and low-level systems work including device drivers and Linux kernel contributions. Track record of measurable impact: 80% voice latency reduction, 5× inference speedup, and 60% cost savings. Also writes technical blogs on Linux, open source, and computer science with 1000+ monthly readers.
Experience
Backend Engineer Intern — Superdash
Bengaluru (Onsite) · June 2025 – Present
- Optimized and deployed edge ML models for voice interruption detection, audio de-noising, and forced word alignment, plus self-hosted TTS (5× speedup, 1000ms → sub-200ms); served via WebAssembly, vLLM, and ONNX Runtime, sustaining sub-100ms inference across concurrent live calls.
- Engineered scalable voice agent backend systems in Go and Python, owning full lifecycle from API design through production deployment; refactored WebSocket layer to cut voice call latency by 80% (500ms → 100ms).
- Designed and shipped an event-driven automation pipeline (akin to n8n) as a core product feature, fully owning the initiative from client co-design and architecture through testing and production rollout; directly drove new client acquisition and revenue growth.
Python Developer Intern — Weya AI
Noida (Remote) · Dec 2024 – Feb 2025
- Built a real-time voice agent backend (FastAPI, WebSockets) scaling to 100+ concurrent connections; profiled 10+ system metrics to diagnose and resolve LLM inference and speech pipeline bottlenecks, improving streaming responsiveness and p99 latency under sustained load.
Summer Research Intern — NIT Jalandhar
Jalandhar, Punjab (Onsite) · May 2024 – July 2024
- Designed and trained neural network models achieving ~95% predictive accuracy via hyperparameter tuning and feature engineering; co-authored a peer-reviewed paper on model architecture and experimental validation, accepted and published in an indexed journal.
Projects
DeadEnv
- Built a cross-platform CLI secrets manager in Go replacing plaintext
.envfiles with OS-native keychain storage (cgo on macOS, libsecret on Linux, WinCred on Windows) behind biometric authentication. - Designed an AES-256-GCM + Argon2id encrypted export format with deliberate wrong-password/tamper error collapsing to prevent oracle attacks; backed by a git audit log and interface-driven architecture enabling full unit and fuzz test coverage.
Spectacle OCR Screenshot ★ 90
- Developed a C++/Qt desktop utility integrating Tesseract OCR and ZXing for low-latency text extraction and QR decoding; achieved native-performance system integration on Linux.
Ricoh SP 200 Linux Driver
- Reverse-engineered the Ricoh SP 200 proprietary print protocol from USB captures of the Windows driver, implementing a native CUPS filter in C using libcups, libcupsimage, and JBIG1 compression (ITU-T T.82).
- Captured and decoded USB traffic between the Windows driver and hardware through systematic binary protocol analysis, reconstructing the full print job lifecycle to enable correct multi-page printing support on a device with no official Linux driver.
Skills
- Languages: Python, Go, TypeScript, C, C++, Rust, Bash
- Frameworks & Tools: FastAPI, Node.js, WebSockets, REST APIs, Qt, Textual, WebAssembly
- Cloud & DevOps: AWS (EC2, S3, SQS, EventBridge), GCP APIs, Docker, GitHub Actions, CI/CD, Linux, Git
- Core Expertise: Distributed Systems, Backend Architecture, Systems Programming, LLM Systems, Machine Learning
Education
JSS Academy of Technical Education — Noida, Uttar Pradesh
B.Tech in Computer Science and Engineering · CGPA: 7.9 · 2022 – 2026
Activities & Achievements
- Technical Contributor, Letscode.in (Feb–May 2025): Authored technical blogs on backend systems, Linux, ML, and GenAI; built and shipped an AI interview simulator and resume analyzer — both deployed and actively used by the platform’s user base.
- Amazon ML Summer School 2024 — Selected from over 1,000,000 applicants nationwide for the competitive machine learning program.
- Linux Kernel Contributor (ALSA/HDA subsystem): Diagnosed a hardware mute LED bug on HP Pavilion Plus by manually probing Realtek ALC245 COEF registers with
hda-verb; submitted a patch merged into mainline Linux (sound/hda/codecs/realtek/alc269.c) adding the missing PCI quirk for this laptop model.