I'm a 2nd-year Software Engineering student at Epitech Nice (Grande École du Numérique, promo 2030), operating at the intersection of systems engineering, cybersecurity, and applied AI.
My work spans deliberately across the stack — from low-level C/C++ firmware on embedded microcontrollers, to secure B2B SaaS architecture, to multi-agent AI-driven quantitative trading systems. I care about building things that actually work in production, with a strong bias toward security-first design (RLS, anti-IDOR, network segmentation) and pragmatic AI integration rather than hype.
Currently completing a 6-month engineering internship at Actis (Telis Monaco), defining MSSP tenant architecture and developing GRC (Governance, Risk & Compliance) tooling in PHP/MySQL, while founding and building OnVoice as a solo technical founder.
- 🎓 Education: Epitech Nice — 2nd year, Programme Grande École
- 🛡️ Focus: Cybersecurity, GRC, systems programming, full-stack & AI-assisted engineering
- 🚀 Founder: OnVoice — B2B SaaS automating invoice recovery for BTP artisans
- 🏆 3x Hackathon builder (Orange Business, PoC x Ramify)
Open To: Software Engineering internships/roles · Cybersecurity & GRC opportunities · AI/ML-driven product engineering · Technical collaborations
Languages
Frontend
Backend & Databases
Cloud, Security & Tooling
| Domain | Proficiency | Details |
|---|---|---|
| Network Security / Pentesting | ⭐⭐⭐⭐☆ | Recon & enumeration with Nmap, brute-forcing with Hydra, web app testing with Burp Suite |
| GRC (Governance, Risk & Compliance) | ⭐⭐⭐⭐☆ | Building an internal GRC/MSSP tenant tool in PHP/MySQL at Actis Monaco |
| Embedded / IoT Security | ⭐⭐⭐☆☆ | ESP32-based passive WiFi IDS detecting deauth/evil-twin attacks in promiscuous mode |
| AI/ML for Decision Systems | ⭐⭐⭐⭐☆ | Random Forest models, local LLM agents (Ollama, Llama 3), LangChain pipelines |
| Secure SaaS Architecture | ⭐⭐⭐⭐☆ | Supabase RLS, anti-IDOR design, Stripe/Twilio/Brevo secure integrations |
💰 crypto_trader_V2 — Multi-Agent Quantitative Trading System
Event-driven microservices platform for automated multi-bot cryptocurrency paper trading (BTC/USDT), running 4 parallel quantitative strategies orchestrated via Docker Compose, with Random Forest ML validation and dual-layer LLM (Llama 3) sentiment analysis.
| Stack | Python, Docker, Random Forest, Llama 3, Supabase, Streamlit |
| Scale | 4 isolated strategy bots (OBI scalper, EMA momentum, AI sentiment oracle, Bollinger reversion) |
| Performance | Async 4-worker architecture, Kelly-criterion position sizing |
| Security | 15% drawdown kill switch, isolated bot containers |
| Impact | End-to-end quant research pipeline, from signal generation to risk-managed execution |
| Repository | github.com/mathish06/crypto_trader_V2 |
A full hedge-fund-style decision engine combining classical technical analysis, machine learning, and LLM-based validation — built solo to explore how multi-agent systems can manage risk in volatile markets.
📡 Opal — AI-Powered SD-WAN Fleet Audit Tool
Built during the Orange Business × Epitech Hackathon 2026 (2nd place). Analyzes real network telemetry across a fleet of 90 VeloCloud routers to recommend a cost-optimized target architecture, with a fully local LLM agent for natural-language querying of the audit database.
| Stack | Python, SQLite, LangChain, Ollama, React, Tailscale |
| Scale | Audit & migration modeling across 90 production routers |
| Performance | Telemetry-driven decision algorithm (throughput, tunnels, flows) |
| Security | 100% local LLM inference, VPN-secured communications (Tailscale) |
| Impact | Identified an architecture generating 64% cost savings on the migration |
| Repository | github.com/mathish06/Opal |
📶 Sentinel-IDS — Portable WiFi Intrusion Detection System
Autonomous Blue Team hardware device built on ESP32, monitoring 802.11 networks in promiscuous mode for real-time packet analysis and passive detection of Deauthentication flooding and Evil Twin attacks via channel hopping.
| Stack | C++, ESP32, PlatformIO, embedded hardware |
| Scale | Real-time single-device monitoring with OLED + RGB feedback |
| Performance | Continuous channel-hopping packet capture in promiscuous mode |
| Security | Passive detection of Deauth flooding and Evil Twin attacks |
| Impact | Fully autonomous, standalone Blue Team monitoring device |
| Repository | github.com/mathish06/Sentinel-IDS |
📊 trading_bot — Quantitative Channel Breakout Strategy
Built during the PoC × Ramify Hackathon. Implements a Channel Breakout strategy using comparative technical analysis and signal normalization over rolling windows, with strict algorithmic portfolio and cash allocation management.
| Stack | Python, quantitative finance, algorithmic trading |
| Scale | Single-strategy portfolio allocation engine |
| Performance | Rolling-window signal normalization |
| Security | Strict algorithmic cash/allocation constraints |
| Impact | Hackathon-validated systematic trading strategy |
| Repository | github.com/mathish06/trading_bot |
🐚 42sh-workspace — Unix Shell Implementation in C
A fully-featured Unix shell written in C as part of Epitech's TEK1 curriculum — implementing piping, redirections, history, tab-completion, aliases, and globbing from scratch.
| Stack | C, POSIX systems programming |
| Scale | Full shell command pipeline |
| Performance | Low-level process & signal management |
| Security | Safe parsing & execution of arbitrary shell input |
| Impact | Core systems-programming foundation project |
| Repository | github.com/mathish06/42sh-workspace |
💼 OnVoice — B2B SaaS for Automated Invoice Recovery
Premium B2B SaaS automating invoice recovery for BTP (construction trades) artisans via a multi-channel autopilot (SMS, Email, registered mail). Full development of the financial dashboard, the cron-based sequencing engine, and secure third-party integrations.
| Stack | FastAPI, Python, Supabase, HTMX, Tailwind, Stripe, Twilio, Brevo |
| Scale | Solo-built, production B2B SaaS for TPE/PME artisans |
| Performance | Automated multi-channel recovery sequencing (Cron-driven) |
| Security | Anti-IDOR design, Supabase Row-Level Security |
| Impact | Founded and operated as an active micro-entreprise (SIREN 101650919) |
| Repository | Private — see portfolio |
July 2026 — December 2026
Cybersecurity & data protection engineering internship. Defining architecture and implementing an MSSP tenant for an internal GRC (Governance, Risk & Compliance) solution.
- Designed and structured the MSSP tenant architecture for the GRC platform
- Developed and modified core features in PHP / MySQL
- Worked within a cybersecurity & compliance-focused engineering team
PHP MySQL GRC Cybersecurity MSSP
2026 — Present
Solo founder and developer of a B2B SaaS automating invoice recovery for BTP artisans (TPE/PME, 1–50 employees).
- Built the full product: financial dashboard, sequencing engine, payment & communication integrations
- Defined go-to-market strategy targeting CAPEB/CCI institutional partnerships
- Operates as an independent micro-entreprise based in Nice
FastAPI Python Supabase Stripe SaaS Architecture
| Recognition | Details |
|---|---|
| 🥈 2nd Place — Orange Business × Epitech Hackathon | SD-WAN fleet audit platform optimizing a 90-router migration with 64% projected cost savings |
| 🏗️ PoC × Ramify Hackathon | Built a quantitative Channel Breakout trading algorithm under time constraints |
Fortinet
current_focus:
learning:
- Advanced network security & GRC frameworks
- Applied machine learning for decision systems
building:
- OnVoice — B2B SaaS invoice recovery platform
- Multi-agent quantitative trading systems
exploring:
- AI-assisted DevSecOps workflows
- Embedded security (ESP32 / IoT)
open_to:
- Software engineering internships & roles
- Cybersecurity / GRC opportunities
- Technical collaborations
"Build systems that work — at a scale that matters."