SM
Shaf MuhammadToronto, Canada
BSc. Computer Science · York University
Cloud-Native & Distributed SOA
Software Engineer @ Intact Financial Corp.

Shaf Muhammad

Full-Stack Software Engineer & Photographer

Designing resilient systems, modern interfaces, and capturing the physical world. Currently engineering cloud-native infrastructure at Intact Financial Corp.

Professional Timeline

Software Engineer

Intact Financial Corp. — Toronto, ON
May 2025 – Present

Collaborated in a cross-functional Agile team to enhance an underwriter platform supporting billions in insured assets.

Designed and deployed loosely coupled RESTful APIs within a distributed Service-Oriented Architecture (SOA), utilizing Java Spring and Angular to improve system modularity and deployment speed.
Leveraged GitHub, Docker, and OpenShift for version control, containerization, and deployment in cloud-native environments.
Accelerated incident resolution across production and dev environments by integrating AI engineering agents (Devin, CoPilot) to automate root cause analysis, debugging, and unit test generation.
Stack:Java SpringAngularDockerOpenShiftGitHubDevin AIGitHub CoPilotRESTful SOAAgile

Software Engineer

Ontario Government — Toronto, ON
May 2024 – Apr 2025 (12 mos)

Developed secure, Java-based applications used across ministries, strictly adhering to government security, accessibility, and governance standards.

Built dynamic, accessible UI components with Angular, ensuring strict compliance with WCAG 2.1 accessibility standards (WAVE, JAWS, NVDA).
Engineered backend RESTful APIs using Spring Boot and established automated CI/CD deployment pipelines via Azure DevOps.
Optimized SQL queries and designed relational database schemas, strictly adhering to government security and governance guidelines.
Stack:AngularJava Spring BootSQLAzure DevOpsWCAG 2.1 AASeleniumPostmanJUnit
01 / 05 — Tap To Dine

Engineering Works

High-Throughput ASGI Event Loop & Distributed POS State Engine

Tap To Dine

Architecture BlueprintSocket.IO / FastAPI (ASGI) · Supabase PostgreSQL (RLS) · Firebase RS256 JWT

A real-time restaurant POS and ordering engine built with FastAPI, WebSockets, and row-level isolated PostgreSQL to coordinate live kitchen and table transactions.

Bi-Directional ASGI WebSocket Multiplexing

Engineered an event-driven Socket.IO broker running atop an asynchronous FastAPI event loop, multicasting table-side mutations to kitchen display terminals with sub-38ms propagation latency. Features exponential backoff reconnection protocols and automatic SSE fallback.

Kernel-Level Postgres RLS & RS256 Verification

Multi-tenant data isolation enforced strictly at the database engine level via Supabase PostgreSQL RLS policies. Custom PL/pgSQL routines parse and validate asymmetric Firebase RS256 JWT claims directly within transactional execution plans.

Topology Schematic · Tap To Dine Cluster
ASGI Event Loop
EDGE CLIENTSAngular / MUX VideoVector Clock Sync
EVENT BROKERFastAPI / Socket.IOUvicorn ASGI Loop
ACID DATA LAYERSupabase PostgreSQLRS256 RLS Isolation
P99 Event Latency< 38 ms
Tenancy ModelKernel-Level RLS
Event LoopFastAPI Uvicorn
Live Application: Production Edge POSVisit taptodine.app
FastAPI (ASGI)Python 3.12Socket.IOSupabase RLSPostgreSQLFirebase RS256Vector ClocksChart.jsMUX HLSRender CI/CD
Distributed Market Data Ingestion & NLP Transformer Pipeline

Novi Finance

Architecture BlueprintNode.js / Python Microservices · FinBERT Sentiment Models · Vercel Edge Cache

A full-stack portfolio analytics engine built with Angular, Node.js, and Python microservices to automate growth modeling and sentiment-driven market insights across TFSA, RRSP, and FHSA accounts.

Asynchronous Message Brokering & Ingestion Queue

Decoupled high-frequency market tick feeds and SEC EDGAR disclosures through an asynchronous Python worker pipeline. In-memory message queues buffer bursts of over 18,000 events/minute, preventing database thread starvation.

FinBERT Transformer Sentiment Inference Pipeline

Containerized microservice running quantized FinBERT NLP models over earnings call transcripts and real-time financial dispatches. Computes token-level sentiment polarity and confidence score tensors in sub-100ms inference cycles.

Topology Schematic · Novi Market Pipeline
FinBERT Inference
TICK & SEC INGESTPython Async Workers18k events/min Queue
NLP VECTORIZERPyTorch / FinBERT98ms Tensor Inference
EDGE COMPUTATIONVercel Edge API94.2% Hit Cache
Inference Latency98 ms (FinBERT)
Edge Hit Ratio94.2% Cache Hit
Ingestion Queue18,000 /min
Live Platform: Distributed Market IntelligenceVisit novi-finance.vercel.app
Node.jsPython 3.12FinBERT / PyTorchRedis QueueWebSocket BusVercel EdgeStale-While-RevalidateFirebase Auth
Hardware-Level POS Handshake Protocol & Polyglot Cloud Archival

Receipto

Architecture BlueprintFlutter / Dart / Kotlin (JNI Channels) · MongoDB / AWS S3 · Dynamic POS QR Cryptography

An enterprise digital receipt infrastructure digitizing physical point-of-sale transactions via a custom dynamic QR handshake protocol, backed by polyglot AWS S3 blob storage and compound-indexed MongoDB replica sets.

Cryptographic POS Dynamic QR Handshake Protocol

Engineered a zero-paper transaction protocol encoding compressed, cryptographically signed POS payload tokens into dynamic terminal QR codes. Client camera sweeps execute mutual HMAC verification and ingest itemized carts in under 250ms without intermediate BLE pairing.

Decoupled Polyglot Persistence: AWS S3 + MongoDB Replica Sets

Asymmetric data tiering architecture: immutable receipt image and PDF binary artifacts are pushed directly to Amazon S3 via pre-signed upload URLs, while normalized transaction line items are indexed in MongoDB with compound text indexes for instant SKU regex lookups.

Topology Schematic · Receipto POS Bridge
HMAC Protocol
TERMINAL PROTOCOLDynamic POS QRHMAC Verification
NATIVE JNI BRIDGEKotlin / CameraXZero-Jank Threads
POLYGLOT STORAGEAWS S3 + MongoDBPre-Signed Object Store
Handshake Verification< 250 ms (HMAC)
Storage ArchitectureS3 Blobs + Mongo Replica
Thread ModelZero-Jank Native Threading
Application Architecture & DemonstrationView Demo Video
Flutter / DartKotlin JNIAndroid CameraXAVFoundationAWS S3 Pre-SignedMongoDB ReplicaHMAC AuthPython Microservices
Zoneless Signal State Engine & Zero-CLS Image Pipeline

Astro Captures

Architecture BlueprintAngular 18 (Signals / Zoneless) · ImageKit / Cloudflare Workers · Statically Tiered JSON

A high-performance photographic asset delivery platform engineered with Angular 18 Signals, Cloudflare Workers edge caching, and an ImageKit dynamic transformation pipeline.

Double-Buffered Zero-CLS Blur-Up Loading & Disk Cache Reuse

Eliminated Cumulative Layout Shift (CLS: 0.00) by reusing pre-cached 800px grid thumbnails as in-memory decoded bitmap placeholders within calculated aspect-ratio bounds. The 2048px display layer is asynchronously pre-decoded via HTMLImageElement.decode() prior to opacity crossfades.

Angular 18 Reactive Signals & Zoneless Node Recycling

Architected state controllers using signal() and computed() primitives to drive view mutations (masonry, carousel, feed) independently of Zone.js monkey-patching. Employs Angular 18 native @for blocks with deterministic identity tracking to eliminate DOM reallocation.

Topology Schematic · Astro Captures CDN
Zero-CLS Pipeline
EDGE WORKER RUNTIMECloudflare Workers< 35ms Global TTFB
IMAGE TRANSCODINGImageKit Edge CDNAVIF / WebP Mipmaps
REACTIVE CLIENT COREAngular 18 Signal GraphZoneless DOM Recycle
Layout Stability0.00 CLS (Double Buffered)
Global Edge TTFB< 35 ms (Cloudflare)
Image DecodeAsync decode() Stack
Production App: Angular 18 StandaloneVisit astrocaptures.com
Angular 18 SignalsZoneless StateCloudflare WorkersImageKit CDNAVIF / WebP TranscodePointerEvent APIZero-CLS PipelineDeep Linking
Low-Latency macOS Keystroke Injection & Context-Aware HUD · Open Source

DevDeck

Architecture BlueprintSwift 5.7+ · AppKit / SwiftUI · CoreGraphics CGEvent Tap · POSIX Process Workers

A high-performance, native macOS developer utility delivering context-aware macros, universal snippet expansion, and sub-millisecond keystroke automation through an intuitive radial HUD.

Low-Latency Keystroke Injection & CGEvent Taps

Engineered a C-level CoreGraphics event tap pipeline via CGEventTapCreate and AXUIElement to intercept global hotkeys with sub-1ms response latency. Simulates hardware-level HID scancodes directly into frontmost application key-down event loops.

Context-Aware Frontmost App Profile Switching

Monitors workspace window states using NSWorkspace.didActivateApplicationNotification, dynamically hot-swapping macro layouts, snippets, and radial toolbars as active focus transitions across Xcode, VS Code, or creative suites.

Topology Schematic · DevDeck Core Engine
Native Swift Engine
SYSTEM EVENT TAPCoreGraphics / CGEvent< 1ms Key Injection
WORKSPACE OBSERVERAppKit / NSWorkspaceDynamic App Profile
SUBPROCESS WORKERSwift Process / ShellPOSIX Async Exec
Input Latency< 1 ms (Native Tap)
Memory Footprint18.4 MB (Zero-Electron)
Target PlatformmacOS 13.0+
Open Source: MIT LicensedGitHub Repository
Swift 5.7macOS 13.0+AppKit / SwiftUICoreGraphics (CGEvent)AXUIElement APINSWorkspace ObserverPOSIX SubprocessesMIT Open Source

The Loadout & Credentials

Infrastructure & Industry Credentials

Education & Industry Credentials

Academic Degree
York University — BSc. CS
Sept 2022 – Aug 2026
Cloud Certification
Microsoft Azure (AZ-900)
Certified Azure Cloud Practitioner
Data Intelligence
Databricks Fundamentals
Certified Lakehouse & Analytics

Frontend Engineering

TypeScriptJavaScriptAngularReactFlutterDartNode.jsHTML/SCSSGSAPTailwind CSSAngular MaterialThymeleafBootstrap

Backend & Systems

PythonJavaSQLSpringBootFastAPIMavenPostgreSQLSupabaseSQLiteMongoDBKotlinSwiftCVerilog HDLRISC-V

Cloud, DevOps & Infra

AzureAWSGCPDockerOpenShiftVercelFirebaseFigmaGitUnix/BashPandasSeleniumPostmanJiraWAVEJAWSNVDA

AI & Agentic Tooling

Devin AIAntigravityClaude CodeCursorCodexGitHub CoPilotRovo

Architectural Disciplines

●Distributed REST & SOA
●WCAG 2.1 AA Compliance
01 / 06 — Glacial Equilibrium

Photography

01/

Glacial Equilibrium

35mm · f/5.0 · 1/250s · ISO 400
Turquoise glacial waters reflecting dense alpine spruce forest wall at Joffre Lakes Provincial Park
EXIF Metadata
35mm · f/5.0 · 1/250s · ISO 400
Canon EOS Rebel SL2·EF-S 18-55mm f/4-5.6 IS STM
Subject: Alpine Glacial Reflection · Coast Mountains, BC
Optics: EF-S 18-55mm f/4-5.6 IS STM · Canon EOS Rebel SL2Landscape Limnology · Aug 18, 2025
02/

Echelon & Celestial Crescent

55mm · f/5.6 · 1/4000s · ISO 400
Three P-51 Mustang fighter planes in tight echelon formation banking across azure sky with crescent moon visible
EXIF Metadata
55mm · f/5.6 · 1/4000s · ISO 400
Canon EOS Rebel SL2·EF-S 18-55mm f/4-5.6 IS STM
Subject: P-51 Mustang Formation · Toronto Waterfront Sky
Optics: EF-S 18-55mm f/4-5.6 IS STM · Canon EOS Rebel SL2Aviation Telephoto · Sept 6, 2026
03/

Architectural Horizon at Dawn

55mm · f/7.1 · 1/400s · ISO 100
Golden hour dawn illumination across rolling green hills towards the silhouette of Mississauga City Centre skyline
EXIF Metadata
55mm · f/7.1 · 1/400s · ISO 100
Canon EOS Rebel SL2·EF-S 55-250mm f/4-5.6 IS
Subject: Mississauga Skyline Silhouette · Golden Hour Contour
Optics: EF-S 55-250mm f/4-5.6 IS · Canon EOS Rebel SL2Architectural Topography · Sept 27, 2026
04/

Downy Woodpecker in Canopy

250mm · f/5.6 · 1/200s · ISO 6400
Downy Woodpecker with vivid red crown patch perched upright on textured tree bark with soft bokeh background
EXIF Metadata
250mm · f/5.6 · 1/200s · ISO 6400
Canon EOS Rebel SL2·EF-S 55-250mm f/4-5.6 IS
Subject: Dryobates pubescens · Riverwood Conservancy, Mississauga
Optics: EF-S 55-250mm f/4-5.6 IS · Canon EOS Rebel SL2Wildlife Ornithology · Sept 22, 2026
05/

Lavender Cosmos

194mm · f/5.6 · 1/640s · ISO 1600
Lavender Cosmos flower in bloom with soft bokeh at Riverwood Conservancy
EXIF Metadata
194mm · f/5.6 · 1/640s · ISO 1600
Canon EOS Rebel SL2·EF-S 55-250mm f/4-5.6 IS
Subject: Cosmos bipinnatus · Riverwood Conservancy, Mississauga
Optics: EF-S 55-250mm f/4-5.6 IS · Canon EOS Rebel SL2Botanical Macro · Sept 22, 2026
06/

Urban Core & Corridor

55mm · f/8.0 · 1/500s · ISO 100
Gardiner Expressway approach toward the downtown Toronto skyline and CN Tower
EXIF Metadata
55mm · f/8.0 · 1/500s · ISO 100
Canon EOS Rebel SL2·EF-S 18-55mm f/4-5.6 IS STM
Subject: Metropolitan Transit Arterial · Toronto Skyline
Optics: EF-S 18-55mm f/4-5.6 IS STM · Canon EOS Rebel SL2Urban Architecture · Sept 6, 2026

Curated visual works and photo essays archived at astrocaptures.com ↗

Visit Astro Captures