ronit@workstation:~$ aero doctor --summary
──────────────────────────────────────────────────────────────────────────
OS :: Aero Linux v1.5.0-Pulsar [Kernel 7.0 / Rolling Edge / Wayland]
Host / CPU :: AMD Ryzen 5 5600H (6C/12T @ 4.2GHz / Low-Latency P-State)
Memory :: 15.3 GB Physical [7.5 GB Dynamic zRAM ZSTD In-Memory Compression]
Education :: CSBS @ Netaji Subhash Engineering College (2023–2027)
Core Stack :: Rust, Java (Loom/Virtual Threads), TypeScript, Node.js, C, PostgreSQL, Redis
Philosophy :: "Zero Bloat • First-Principles Depth • Hand-Crafted Code"
──────────────────────────────────────────────────────────────────────────- ⚡ Aero-Linux — Ultra-Lean Developer OS & Native Systems Suite. Sub-350MB idle footprint, dynamic zRAM (ZSTD) memory optimization, 59 native zero-terminal GTK3 developer applications, 133/133 unit tests, eBPF kernel tracing, ephemeral KVM microVMs, automated Debian package builds (
.deb), and live web portal ↗. - ⚡ FinFlow-Core — Enterprise Banking & Financial Aggregation REST Platform. [Live API ↗] Built on Spring Boot 3.3, Java 17, Spring Security 6 (stateless JWT authentication), PostgreSQL 16, and Docker. Features multi-tenant isolation, automated scheduled market data ingestion, custom JPQL financial aggregation, and an offline UPI mesh settlement engine (indexed on XERJ Corpus Hub ↗).
- 🎙️ Vox-Intel-Engine — Multi-Modal Speech Telemetry & Acoustic DSP AI Engine. Built with Python 3.11 and FastAPI for real-time speech analytics. Extracts pitch fundamental frequencies via autocorrelation DSP, spectral energy distribution via FFT, and lexical delivery metrics (WPM, filler word density, type-token ratio).
- ⚡ Dine-Luxe — High-Concurrency POS & Reservation Engine. [Live on Vercel ↗] Real-time table lifecycle management, optimistic locking against double-booking, sub-millisecond cache invalidation, and transactional consistency built with TypeScript, Node.js, Express, Prisma ORM, and PostgreSQL.
- ⚡ BurnRate — Terminal HUD, Hard Spend Caps & Cache Guard for Coding Agents. Never wake up to surprise $40 Claude Code or Cursor bills. Pure-Rust zero-overhead reverse proxy enforcing hard budget circuit breakers (
HTTP 402), byte-level prompt-cache invalidation linting, token-cost optimal cascade routing, multi-agent quotas, sliding-window velocity brakes, native MCP server, and Prometheus export. Published on Crates.io (burnrate-agent). - ⚡ Agent-2PC — Heterogeneous Two-Phase Commit (2PC) & WAL Engine for AI Agents. Eliminates dirty repository state, broken syntax, and hallucinated rollbacks across autonomous agent tool pipelines. Rust engine featuring crash-consistent binary Write-Ahead Log (
fsync+ CRC32), isolated scratchspaces, deterministic rollbacks, and drop-in Model Context Protocol (MCP) server support. - ⚡ Agent-MemGC — The Linter, Cache-Layout Optimizer & Generational Garbage Collector for AI Agent Instructions. Audit, clean, and deduplicate
CLAUDE.md,AGENTS.md, and.cursorrules. Detects dead file path references, resolves conflicting directives, eliminates prompt-cache-busting volatile prefixes, and auto-fixes context bloat in <15ms. Native MCP server published on Crates.io (agent-memgc). - ⚡ Local-Jev — 100% Offline, Zero-Cost Drop-In Alternative to TypeSafe Jev System One Decisions. Sub-18ms local CPU decision engine (
Choice,Score,Noul,Route). Single-forward-pass logit extraction, SIMD quantized linear layer, speculative cascade routing, contextual bandit feedback, zero API keys, and seamless Axum REST server published on Crates.io.
- ⚡ AvioStream — Closed-Loop Autonomous Fleet Telemetry & Predictive Maintenance Platform. Built in Java 21 Loom (>48,000 msgs/s virtual thread ingestion buffer) with embedded ONNX runtime edge inference (0.014ms), dual ML (Isolation Forest + LightGBM asymmetric NASA PHM loss), and sub-15ms deterministic SAP PM / IBM Maximo CMMS work-order dispatch.
- ⚡ Apex-Matching-Engine — Ultra-Low Latency Continuous Double Auction Order Book. Price-time priority (FIFO) execution engine built in Java 21 Loom, delivering in-memory L2 market depth updates, concurrent order matching, and real-time STOMP WebSocket feeds exceeding 50,000 ops/sec.
- 🇮🇳 Bharat-Spatial-Engine — National-Scale Geospatial Hierarchy & Proximity Query Engine. Sub-millisecond reverse geocoding and radius proximity search across India's 650,000+ census settlements and MDDS codes. 2D geographic KD-Tree with spherical Haversine pruning and transliteration-tolerant Indic trigram search, clocking 38,400+ QPS with 2.09ms p99 latency in Java 21 Loom.
- ⚡ Vectra-Core — In-Memory Vector Search Engine & HNSW Graph Index. High-dimensional approximate nearest neighbor (ANN) retrieval in Java 21 Loom virtual threads. 8-bit scalar quantization (SQ8) with 75% RAM reduction, filtered hybrid vector search, and verified 16,900+ QPS with 0.28ms median latency.
"Simplicity is prerequisite for reliability." — Edsger W. Dijkstra