.png)
★
Omair Afzal.
Active open-source contributor to Prisma, NestJS, Qdrant, LanceDB, Tauri, and 20+ other projects across the Rust, TypeScript, and Python ecosystems. Currently leading engineering at ClickGUARD and architecting AI-driven products for early-stage startups. Specializes in event-driven architectures, vector-search infrastructure, and building resilient systems that handle millions of daily transactions.
What they do best.
AI & Intelligent Systems
- Retrieval-Augmented Generation (RAG)
- Vector Search & Embeddings
- Multi-Agent Orchestration
- LLM Integration & Prompt Engineering
- Cognitive Architecture Design
- Knowledge Graph Construction
Open Source & Developer Tooling
- Prisma ORM Contributor
- NestJS Core Contributor
- Rust Ecosystem (Qdrant, LanceDB, Tauri)
- Symfony Framework
- Technical Writing & Documentation
- Community Mentorship
Systems Architecture
- Distributed Systems Design
- Event-Driven & CQRS Patterns
- Vector Databases (Qdrant, Milvus, LanceDB)
- High-Performance Rust Services
- Microservices & Domain-Driven Design
- Graph Databases (Neo4j, ArangoDB)
24 projects, shipping consistently.
Code shipped to open-source projects real teams depend on.
OpenClaw
2024 — presentShipped core fixes to the OpenClaw agentic framework across routing, auth, and platform integrations. Merged gateway auth-token validation fix and opened multiple PRs improving CLI reliability, WhatsApp/Telegram/Signal channel modules, and Linux compatibility.
Prisma ORM
2019 — presentResolved driver-adapter edge cases in the Prisma ORM monorepo, including MSSQL connection-string escaping (#29158). Opened additional PRs for MariaDB special-character handling and cross-realm Uint8Array compatibility in jsdom environments.
NestJS
2017 — presentStrengthened test coverage across the NestJS framework sample applications. Submitted unit and e2e tests for the serializer and serve-static samples, improving reliability of official example projects used by thousands of developers.
Keep
2023 — presentBuilt the Nagios monitoring provider for Keep, a YC W23-backed open-source alert management platform. Implemented bidirectional alert ingestion and webhook-based real-time notifications.
Archestra
2024 — presentImplemented Groq and DeepSeek LLM provider integrations for Archestra, a cloud-native MCP registry and gateway. Expanded the platform's multi-model AI capabilities.
ComfyUI
2023 — presentFixed scroll-event propagation leaks in ComfyUI's TextPreviewWidget, resolving canvas interaction instability in the leading open-source AI image-generation interface.
Cal.com
2021 — presentDelivered multiple features to Cal.com, the open-source scheduling platform — booking question field types for routing forms, guest availability checks on reschedule, and email-invite functionality for team assignments.
Coolify
2021 — presentAdded SurrealDB service templates to Coolify, the open-source self-hosting platform. Created both standalone RocksDB and distributed TiKV configurations.
Documenso
2023 — presentFixed a UI overflow bug in Documenso, the open-source DocuSign alternative — restoring full settings-page functionality on the public profile template.
tscircuit
2023 — presentFixed STEP-file generation in tscircuit where component rectangles were incorrectly merged into a single invalid solid — producing valid CAD exports for hardware design workflows.
Nuclei
2020 — presentIntegrated the typos spell-checker into Nuclei's CI pipeline and corrected misspellings across 15 files for ProjectDiscovery's widely used vulnerability scanner.
Cap
2024 — presentBuilt deeplink actions in Rust and a full Raycast extension for Cap, the open-source screen-recording tool — enabling control of recordings, camera, and microphone from the macOS launcher.
Firecrawl
2024 — presentAuthored comprehensive MCP server installation guides for Firecrawl covering 30+ environments — Cursor, VS Code, Claude Desktop — reducing onboarding friction.
DN Institute
2023 — presentEnhanced ML reporter prompts for improved market analysis at DN Institute, a crypto market-health research platform. Added integration tests for the similarity-search API.
Qdrant
2021 — presentAdded per-collection Prometheus metrics to Qdrant, the high-performance Rust vector database. Enabled granular monitoring of REST request counts, durations, and response status per collection.
LanceDB
2023 — presentFixed a crash in LanceDB's hybrid search when hard filtering returned empty results. Refactored the 5,000-line table.rs by extracting the merge_insert module.
Foundry
2021 — presentFixed a bug in Foundry where the verify flag suppressed console-log output during script simulation. Restored developer debugging experience for the Rust-based Ethereum toolkit.
Screenpipe
2024 — presentOverhauled the OpenAPI specification for Screenpipe — added 18 missing endpoints, removed 12 stale ones, fixed 15 schema mismatches. Aligned API documentation with the actual codebase.
uv
2024 — presentImproved developer experience in uv, Astral's Rust-based Python package manager. Added debug logging for config-file resolution and clarified ambiguous error messages.
Ruff
2022 — presentExtended Ruff's TD003 linting rule to recognize Jira-style issue IDs (e.g., JIRA-123) in TODO comments alongside GitHub-style links.
Meilisearch
2018 — presentFixed a bug in Meilisearch where attributesToSearchOn errored on empty indexes instead of returning empty results. Improved graceful handling for the Rust-powered open-source search engine.
Tauri
2019 — presentAdded the Rust equivalent of convertFileSrc() to Tauri, enabling server-side conversion of local file paths to asset-protocol URLs. Closed a parity gap between Rust and JavaScript APIs.
Typst
2023 — presentFixed the Typst compiler to correctly infer dictionary types from spread operands in empty collection literals. Resolved a type-inference edge case in the Rust-based typesetting system.
Symfony
2015 — 2020Contributed to the Symfony PHP framework over five years, focusing on HTTP Kernel and Console component improvements. Helped shape the developer experience for one of the most widely adopted PHP frameworks.
Where they've shipped.
Head of Engineering · ClickGUARD
Leading engineering for an ad-fraud protection platform processing millions of daily transactions. Driving technical strategy, team growth, and integration of AI-powered threat detection.
- ◆ Built an autonomous fraud-detection system processing 10M+ daily transactions with sub-second latency
- ◆ Reduced false-positive rates by 40% through vector-similarity models and behavioral analysis
- ◆ Implemented temporal workflow orchestration for complex multi-step threat mitigation
- ◆ Optimized cloud-infrastructure costs by 60% through intelligent autoscaling
Senior Software Architect (Consultant) · Waffle.coach
Architecting the technical foundation for an AI-driven parenting assistant. Designing RAG pipelines, semantic memory systems, and privacy-first data architectures.
- ◆ Built a context-aware RAG system achieving sub-100ms latency across 100M+ vectors
- ◆ Designed a developmental-state tracking module for personalized child-development insights
- ◆ Implemented a hybrid edge-cloud architecture for real-time mobile responsiveness
- ◆ Achieved GDPR compliance through federated learning and on-device processing
Technical Architect (Consultant) · Core Direction
Designing event-driven architectures for high-throughput health-data processing and real-time analytics. Leading migration from monolith to event-sourced microservices.
- ◆ Migrated legacy monolith to event-driven microservices, reducing deployment cycles by 75%
- ◆ Implemented CQRS patterns that reduced read latency by 80%
- ◆ Deployed a real-time leaderboard system handling millions of updates per minute
- ◆ Integrated wearable-device data streams with sub-second end-to-end processing
Senior Software Engineer · AlayaCare
Spearheaded migration from monolithic PHP to a microservices architecture serving enterprise healthcare clients. Built core scheduling infrastructure and GraphQL federation at scale.
- ◆ Designed the core Scheduling Microservice, reducing conflict-calculation time by 90%
- ◆ Built a real-time patient-monitoring dashboard using GraphQL Subscriptions
- ◆ Standardized API design patterns across 20+ microservices serving enterprise clients
- ◆ Introduced mutation testing to CI, measurably increasing test-suite reliability
What they've built.
Autonomous Enterprise Knowledge Graph
A self-organizing knowledge base using autonomous agents to crawl, classify, and link enterprise data sources. Multi-agent orchestration for continuous graph updates, consistency verification, and gap detection.
- Self-healing knowledge graph with automatic inconsistency detection
- Autonomous research agents that fill knowledge gaps proactively
- Hybrid vector-graph retrieval enabling multi-hop reasoning queries
- Enterprise-grade RBAC with granular collection-level permissioning
OpenClaw Execution Engine
The core execution engine enabling AI agents to interact with modern web applications reliably. Handles complex DOM interactions, auth flows, and dynamic content extraction.
- Headless browser orchestration with anti-detection and fingerprint randomization
- Semantic DOM understanding for resilient, self-healing selector generation
- High-performance async runtime built on Rust and Tokio
- Plugin architecture supporting custom action handlers and middleware
Distributed Vector Search Platform
A high-performance vector search platform engineered for billion-scale embeddings. Custom HNSW indexing with distributed sharding for sub-millisecond query latency.
- Distributed index sharding with automatic rebalancing and failover
- Real-time vector insertion with tunable consistency guarantees
- Multi-modal embedding support across text, image, and audio
- Production-grade observability with per-collection Prometheus metrics
Autonomous Fraud Prevention System
A real-time ad-fraud detection system using autonomous agents to simulate user behavior and validate traffic sources. Processes 100K+ events per second.
- Real-time behavioral analysis using deep-learning classification models
- Autonomous traffic-validation agents simulating genuine user patterns
- High-throughput event ingestion handling 100K+ events per second
- Automated rule generation that adapts to emerging fraud patterns