Work Experience of Mohd Faraz

Building and scaling modules across KareXpert's Unified Digital Healthcare Platform — from AI-powered workflow automation to clinical reporting engines and secure full-stack integrations.

• ✔Architected a DAG-based agentic AI workflow engine in Python, implementing nested for-each execution, tool-call observability, and LLM prompt context scaling (20k → 50k chars) to power intelligent automation pipelines.
• ✔Engineered Java/Spring backend services using a decorator pattern for RBAC, data onboarding pipelines, async ticket triage with CompletableFuture, and robust CRM lead deduplication with merge logic.
• ✔Built and maintained Angular/TypeScript frontends including a visual workflow builder, AI chat embed widget, and session-hardened auth flows with JWT binding and CORS allowlist enforcement.
• ✔Diagnosed and resolved critical PDF reporting engine bugs — fixing chart palette overflows, cumulative print banner context freezing, and CSS/font sanitization for GST invoice generation.
• ✔Packaged and shipped internal Node.js dev utilities as cross-platform standalone binaries using pkg, eliminating environment setup friction across Windows, Linux, and macOS.

Co-founded and built one of the most widely used custom recovery interfaces for Android devices worldwide.

• ✔Developed high-security C++/Go flashing utilities, creating automated installation pipelines for custom ROMs and secure updates.
• ✔Ported and compiled recovery builds for dozens of OEM devices, resolving low-level kernel bugs and optimizing partition handling.
• ✔Created sleek custom UI themes that dramatically improved usability compared to standard TWRP recovery boards.