| Research Vector | Focus / Scope | Operational State | Current Target |
|---|---|---|---|
| Human-Centered AI & Usability | Interaction dynamics, AI usability evaluation in ICT systems | [ACTIVE] |
Research group under Prof. Ganesh Bhutkar; scheduled CHIuXD 2026 presentation |
| Video Management Systems (VMS) | VMS architectures, streaming pipelines & camera telemetry | [ACTIVE] |
Structured literature-review corpus of 60 VMS-related academic papers (2015–2025) |
| Edge AI & Video Analytics | Real-time perceptual inference, pipeline latency & edge compute | [ACTIVE] |
Multi-capability analytics verification & RTSP telemetry pipelines |
| Android Transport & Networking | Cellular queue latency, bufferbloat characterization & mobile telemetry | [EXPERIMENTAL] |
Android transport measurement tooling & latency profiling |
| LLM Architectures & Agent Systems | Retrieval-Augmented Generation, tool execution & structured inference | [EXPERIMENTAL] |
Grounded agent workflows, tool execution & document synthesis |
A curated selection of publicly observable systems, technical showcases, and experimental projects.
| System / Repository | Primary Domain | Problem Explored | Technical Stack | Experimental Status |
|---|---|---|---|---|
5G-Bufferbloat-App |
Android Networking / Transport | Investigating queue delay, latency inflation under loaded cellular connections, and transport telemetry on mobile devices | Kotlin Java Android SDK Networking Telemetry |
EXPERIMENTAL (Active exploration) |
ai-video-analytics-showcase |
AI Video Analytics / VMS | Interactive research catalog detailing 28 computer vision capabilities with verified visual demonstrations and operational constraints | Computer Vision Video Analytics VMS Architectures Web UI |
STABLE SHOWCASE (Verified demos) |
pmc-cctv-surveillance-editorial |
CCTV / Surveillance Systems | Interactive technical showcase exploring CCTV infrastructure, AI video analytics and VMS architecture in a municipal-scale scenario | CCTV Systems VMS Architectures Video Analytics Technical Editorial |
ACTIVE SHOWCASE (Technical showcase) |
5G-India-Gaming-Guide— Practical systems guide on optimizing Android 5G tethering / hotspot transport and reducing packet routing jitter for Windows endpoints.
Modern systems engineering and applied AI research require rigorous distinction between configuration, assumption, and measurable reality.
- Measure Before Claiming: No optimization or algorithmic improvement is accepted without baseline and post-intervention comparative instrumentation.
- Distinguish Configuration from Behavior: Parameterizing a pipeline does not guarantee runtime compliance; observable execution telemetry provides the ground truth.
- Preserve Negative Results: Null hypotheses, failed latency thresholds, and unviable inference topologies are preserved as valid engineering signal.
- Defensible Limits: State explicit boundaries, hardware constraints, and failure modes rather than claiming universal generalization.
- Reproducible Tooling: Maintain reproducible configurations, explicit seeds, and controlled environments wherever feasible.
- Institution: Vishwakarma Institute of Technology (VIT Pune)
- Role: Research Intern under Prof. Ganesh Bhutkar
- Research Group Focus: Usability of AI ICT Applications
- Scheduled Presentation: CHIuXD 2026 (Indonesia, December 2026) — Research work examining usability frameworks and human-interaction metrics in AI-assisted ICT systems.
- Corpus Synthesis: Structured literature-review corpus of 60 VMS-related academic papers (2015–2025) examining Video Management Systems, distributed video pipelines, and surveillance analytics architectures.
[ PUBLICATIONS // VERIFIED ENTRIES BEING INDEXED ]
------------------------------------------------------------------------------------------------------
Verified publication entries, citation metadata, and preprint indices are currently being prepared
for public listing.
------------------------------------------------------------------------------------------------------
(Verified entries will be indexed here with venue, year, DOI, and preprint links.)
| Property | Details |
|---|---|
| Title | Personalized Voice Control Car |
| Field | Electronics |
| Filing Date | 11 December 2025 |
| Publication Date | 23 January 2026 |
| International Patent Classification (IPC) | G10L 17/00 · G10L 17/22 · G10L 15/22 · G10L 17/04 · G10L 17/02 |
| Status | Published (Public Record) |
View Detailed Technology Index
- Languages: Python, C, C++, Java, Kotlin, TypeScript, JavaScript, SQL, POSIX Shell
- AI & Perceptual Systems: Computer Vision, OpenCV, Video Analytics, LLM Architectures, RAG Systems, LangChain, Model Inference APIs
- Systems & Infrastructure: Android SDK / NDK, JNI Interfacing, Linux Runtimes, Docker Containers, CMake, Git, GitHub Actions CI/CD
- Web & Interface Engineering: React, Next.js, Node.js, Express, FastAPI, MongoDB, MySQL, RESTful Services, WebSocket Streaming
Ongoing research efforts, active experimental sandboxes, and upcoming public technical repositories:
- 🔬 Cellular Bufferbloat & Transport Feasibility: Instrumenting queue depth, round-trip time variance, and downlink/uplink queuing delays on Android under variable radio link conditions.
- 📹 VMS & Edge Video Analytics Experimentation: Benchmarking RTSP ingest pipelines, multi-camera decode overhead, and low-latency edge inference setups.
- 👥 Human-AI Interaction Usability Frameworks: Developing quantitative usability rubrics for AI-augmented ICT tools in preparation for academic dissemination.
├── Frigate-VMS-Lab [IN DEVELOPMENT // VMS integration & NVR telemetry]
├── VMS-Literature-Evidence-Lab [IN DEVELOPMENT // 60-paper corpus taxonomy & comparative matrix]
├── Human-AI-Usability-Lab [IN DEVELOPMENT // Evaluation scripts & HCI usability rubrics]
└── OpenCV-RTSP-Latency-Lab [IN DEVELOPMENT // Frame-dropping & end-to-end latency benchmarks]
(Repositories will transition to public links upon completion of initial baseline benchmarks and documentation.)
[ RESEARCH / ENGINEERING CONTACT ] · Pune, Maharashtra, India · UTC+05:30 IST
Portfolio Website & Academic Dissemination Links Available Post-Release
Repository telemetry tracked automatically via refresh-profile.yml · Structured profile records maintained in profile.json