Skip to main content
SYS_INIT // DWARKESH.LAB
DWARKESHRAMANI
Computer Engineering · Full-Stack Engineer
INITIALIZING DWARKESH // BUILD SYSTEM...12%
CURRENTLY BUILDING · AI SYSTEMS · DEVELOPER TOOLS · FULL-STACK PRODUCTS
DWARKESH RAMANI // SYSTEMS & AI

EXPLORE WORK
GITHUB / DAKU3011
++++
OPERATOR // IDENT
ID: DAKU3011
Dwarkesh Ramani — Engineer
SYSTEM_ONLINE
UTC+05:30
DWARKESH RAMANICOMPUTER ENGINEER
Surat, India
Full-Stack · AI
MODE: ACTIVE IN THE LAB● READY
FULL-STACK ENGINEER · AI SYSTEMS · DEVELOPER TOOLS · REACT NATIVE · FASTAPI · NEXT.JS 15 · POSTGRESQL · REDIS · DOCKER · GEMINI 2.5 · CHROMADB · SURAT, INDIA · FULL-STACK ENGINEER · AI SYSTEMS · DEVELOPER TOOLS · REACT NATIVE · FASTAPI · NEXT.JS 15 · POSTGRESQL · REDIS · DOCKER · GEMINI 2.5 · CHROMADB · SURAT, INDIA · FULL-STACK ENGINEER · AI SYSTEMS · DEVELOPER TOOLS · REACT NATIVE · FASTAPI · NEXT.JS 15 · POSTGRESQL · REDIS · DOCKER · GEMINI 2.5 · CHROMADB · SURAT, INDIA · FULL-STACK ENGINEER · AI SYSTEMS · DEVELOPER TOOLS · REACT NATIVE · FASTAPI · NEXT.JS 15 · POSTGRESQL · REDIS · DOCKER · GEMINI 2.5 · CHROMADB · SURAT, INDIA ·
01/

SELECTED WORK

FOCUS: MINICODE
01/Full-Stack Engineer & System Architect·SHIPPED

MiniCode

Competitive programming, rebuilt for real engineers.

Next.js 15FastAPIPostgreSQLRedisGoogle GeminiDockerGitHub REST APITailwind CSS
VIEW CASE STUDY
REPOSITORY
LIVE DEMO
SYSTEM TOPOLOGY & EXECUTION PIPELINE6 ARCHITECTURAL NODES
Next.js UI
React 19 / Modern HUD
FastAPI Engine
Async REST API & Webhooks
PostgreSQL
Auth, XP, Milestones
Redis & Celery
Event Queue & Rate Limiter
Docker Sandbox
Isolated Code Execution
Gemini Evaluator
Semantic Code Feedback
EXECUTION FLOW
01 →Student pushes commit to private challenge repository
02 →GitHub Webhook triggers FastAPI ingestion endpoint with commit SHA
03 →FastAPI dispatches evaluation task to Celery / Redis queue
04 →Docker container spins up in a hardened sandbox to execute unit test suite
05 →Gemini 2.5 Flash reviews AST, algorithmic complexity, and idiomatic style
06 →Consolidated test telemetry and AI mentorship feedback push live to Next.js frontend
TRADE-OFF // Git Integration vs In-Browser Editor
Dual-mode GitHub webhook ingestion with fallback sandbox

Fosters industry-standard Git workflow habits (branches, commit messages, PRs) rather than treating code as ephemeral text snippets.

02/AI Engineer & Backend Architect·SHIPPED

AI Hackathon Judge

Multi-persona autonomous consensus judge and multimodal project evaluator.

Python 3.11FastAPIGoogle Gemini 2.5OpenCVAsyncIODockeryt-dlpPydantic
VIEW CASE STUDY
REPOSITORY
LIVE DEMO
SYSTEM TOPOLOGY & EXECUTION PIPELINE6 ARCHITECTURAL NODES
FastAPI Ingestion
Async multi-source collector
BFS Tree Parser
LOC & Dependency Graph
Secret Sweeper
Regex Credential Auditor
Video Pipeline
yt-dlp / Gemini Vision 480p
5-Judge Panel
VC, CTO, PM, UI/UX, Professor
Score Matrix
Consensus Aggregation
EXECUTION FLOW
01 →User submits repository link, live demo URL, pitch deck (PDF/PPTX), and demo video
02 →BFS file crawler parses repository dependency manifests and computes entry point LOC
03 →Security regex engine sweeps tree for exposed AWS keys, GitHub tokens, and private keys
04 →4-layer video pipeline extracts audio transcripts via client-side fetch, cookies, or Gemini Vision
05 →asyncio.gather executes 5 persona prompts in parallel against unified project context
06 →Mathematical aggregator computes Win Probability™ and generates brutal 'Why You Won't Win' critique
TRADE-OFF // Multi-Judge Modeling
5 distinct domain personas rather than a single prompt

A single prompt produces homogenized scores. Distinct personas unmask conflicting trade-offs (e.g., CTO scores high for code, VC scores low for TAM).

03/Mobile & Systems Developer·SHIPPED

GitRemote

Remote Git operations and diff inspection from mobile devices.

React NativeExpoTypeScriptNode.jsExpressNative Git CLILocal Subnet Discovery
VIEW CASE STUDY
REPOSITORY
SYSTEM TOPOLOGY & EXECUTION PIPELINE5 ARCHITECTURAL NODES
Mobile Client
React Native / Expo / TS
Subnet Scanner
Zero-config LAN Discovery
PC Daemon
Express / Node.js Service
Native Git Engine
Host Git & SSH Credentials
Cloud Fallback
GitHub REST API Mode
EXECUTION FLOW
01 →PC daemon starts, binds to local port, and broadcasts presence or responds to subnet probe
02 →Mobile app scans local /24 subnet to auto-pair with the PC daemon without manual IP entry
03 →Daemon executes non-blocking native git status and diff commands against configured repo roots
04 →Mobile client renders colored line additions, deletions, and hunks with horizontal scroll support
05 →User selects hunks to stage, inputs commit message, and triggers authenticated git push
TRADE-OFF // Direct Host Git vs Re-implementing Git in JS
Spawn native host Git CLI via Node.js child_process

Leverages the host's existing SSH keys, GPG signing configs, submodules, and credential helpers without storing private keys in the app.

04/Full-Stack AI Engineer·SHIPPED

Class Intelligence System

Departmental RAG system with ChromaDB and Gemini citations.

FastAPINext.jsChromaDBGoogle Gemini 1.5 FlashBcrypt + SHA-256Tailwind CSSPython 3.11
VIEW CASE STUDY
REPOSITORY
SYSTEM TOPOLOGY & EXECUTION PIPELINE4 ARCHITECTURAL NODES
Next.js Portal
Student & Faculty Interface
FastAPI Server
Document Pipeline & Auth
ChromaDB
Vector Store & Embeddings
Gemini 1.5 Flash
Contextual QA Synthesizer
EXECUTION FLOW
01 →Faculty upload course syllabi, lecture slides, and past papers via authenticated portal
02 →Document parser extracts text, sanitizes formatting, and splits into semantic chunks
03 →ChromaDB indexes vector embeddings using domain-tuned similarity thresholds
04 →Student queries trigger hybrid vector retrieval to fetch top-k relevant source paragraphs
05 →Gemini 1.5 Flash synthesizes direct answer with exact slide / page citation badges
TRADE-OFF // Vector Store Selection
ChromaDB for embedded local persistence

Enables straightforward deployment without expensive managed vector database overhead for department-scale corpora.

05/Lead Backend & AI Engineer·SHIPPED

HackerRank Orchestrate

Visual evidence verification for automated damage claims.

Python 3.11Multimodal AIComputer VisionPydantic Schema ValidationPyTestAutomated Evaluation
VIEW CASE STUDY
REPOSITORY
SYSTEM TOPOLOGY & EXECUTION PIPELINE4 ARCHITECTURAL NODES
Claim Ingestor
Schema Validation Layer
Vision Evaluator
Object & Damage Classifier
Claim Cross-Matcher
Consistency Engine
Audit Logger
Deterministic Verdict
EXECUTION FLOW
01 →Inbound claim payload with conversational logs, user metadata, and visual attachments
02 →Schema validator validates payload integrity against strict competition constraints
03 →Multimodal vision model evaluates visual damage specifics against claimed object type
04 →Cross-verification logic checks consistency with historical claims and conversation transcripts
05 →Decision output formatted according to strict benchmark schema with audit trail
TRADE-OFF // Decision Schema
Strict Pydantic models with automated test suite

Guaranteed zero schema mismatches during the automated 24-hour evaluation harness runs.

02/

DAILY QUOTES

DAILY SYSTEM RECORD

Timeless mental models and wisdom from the pioneers of computing systems.

// THEORY & ALGORITHMS
RECORD #01 / 10CIRCA 1970

“Computer science is no more about computers than astronomy is about telescopes.”

Edsger W. Dijkstra
Turing Award Laureate · Structured Programming Pioneer
03/

EXPERIENCE

TIMELINE

Technical leadership, engineering roles, and developer community stewardship.

2026CURRENT

Backend Engineer Intern

Ornitech Solution

Engineering autonomous AI agents, backend automation workflows, and high-reliability full-stack systems.

AI AgentsAutomation WorkflowsFull-Stack SystemsAPI Engineering
2025 — PresentCURRENT

Student Lead

SCET Coding Club

Technical leadership, mentoring junior developers, organizing algorithmic sprints, and coordinating hackathons across the university.

Technical LeadershipDeveloper CommunityHackathonsMentorship
2025 — PresentCURRENT

Tech Lead

AWS Student Builder Group

Guiding cloud architecture, serverless systems, and AI developer ecosystem workshops for engineering students.

Cloud ArchitectureAWS SolutionsAI Developer ToolsWorkshops
2025

Campus Ambassador

E-Cell IIT Bombay

Driving entrepreneurship initiatives, hackathon participation, and technical startup outreach on campus.

Startup EcosystemTechnical OutreachInnovation Initiatives
04/

LAB

ACTIVE R&D

Things I'm currently figuring out.

AI & AGENTSSTATUS: ACTIVE EXPERIMENT

Agentic Honeypot

Autonomous scam engagement & counter-intelligence extraction engine.

An AI agent honeypot API that intercepts scam messages, maintains plausible multi-turn conversations with malicious actors, extracts IOC intelligence (crypto wallets, phishing links, phone numbers), and reports structured forensics.

JavaScriptLLM Multi-Turn Prompting
EDGE & CVSTATUS: RESEARCH NOTE

Real-Time ISL-to-Speech AR

Edge-AI wearable converting Indian Sign Language gestures to audio.

JavaScriptMediaPipe
SYSTEMSSTATUS: PROTOTYPE

Auto Attendance System

High-throughput facial recognition with ArcFace & YuNet on ASP.NET.

Classroom attendance system capturing high-resolution photos, detecting multi-face clusters simultaneously via YuNet, computing cosine embeddings with ArcFace on ONNX Runtime, and pushing instant updates over SignalR.

C#ASP.NET Core 8
SECURITYSTATUS: OPEN SOURCE

Agent Governance Toolkit

Policy enforcement & zero-trust identity for autonomous AI agents.

PythonTypeScript
ANALYTICSSTATUS: SHIPPED

University Analytics Platform

Centralized multi-college activity aggregation & automated reporting.

Cloud portal aggregating monthly media activities across multiple colleges into unified analytics dashboards with automated AI-assisted quarterly report generation.

HTML/PythonData Extraction
SYSTEMSSTATUS: PROTOTYPE

shivStock Analytics

Real-time stock portfolio and financial metric visualizer.

TypeScriptNext.js
05/

SYSTEM

RUNTIME MATRIX

Architectural matrix verified across production repositories.

FRONTEND6 TECHNOLOGIES
  • ▪React 19
  • ▪Next.js 15
  • ▪TypeScript
  • ▪Tailwind CSS
  • ▪Three.js / WebGL
  • ▪State & Context
BACKEND6 TECHNOLOGIES
  • ▪FastAPI (Python)
  • ▪Node.js / Express
  • ▪REST Architecture
  • ▪Webhooks & IPC
  • ▪AsyncIO
  • ▪Celery Queues
DATA & STATE5 TECHNOLOGIES
  • ▪PostgreSQL
  • ▪Redis
  • ▪ChromaDB (Vector)
  • ▪SQL Optimization
  • ▪Relational Models
AI & MULTIMODAL6 TECHNOLOGIES
  • ▪Gemini 2.5 / 1.5
  • ▪RAG Pipelines
  • ▪Autonomous Agents
  • ▪Multi-Persona Panels
  • ▪Computer Vision
  • ▪MediaPipe / OpenCV
INFRASTRUCTURE5 TECHNOLOGIES
  • ▪Docker Sandboxing
  • ▪AWS Cloud
  • ▪Vercel Edge
  • ▪Render Deployments
  • ▪Linux / POSIX
MOBILE & EDGE5 TECHNOLOGIES
  • ▪React Native
  • ▪Expo
  • ▪Local Subnet Sync
  • ▪UDP / Socket Pairing
  • ▪Edge Inference (ONNX)
06/

ABOUT

ENGINEER

I BUILD THINGS
THAT WORK.

Computer engineering student based in Surat. Moving between product development, multimodal AI systems, backend architectures, and experiments that started as “what if we automated this?”

BASED INSurat, India
STUDYINGComputer Engineering
FOCUSFull-Stack · AI Systems · Developer Tools
CURRENTLYBuilding, experimenting, shipping
07/

CONTACT

REACH OUT

HAVE SOMETHING INTERESTING?

→ rdwarkesh1300@gmail.com
GITHUB / DAKU3011
LINKEDIN / RAMANIDWARKESH