محمد Osman

• Designed Python-based microservices using Celery for distributed task queues, orchestrating 10k+ daily AI inference
requests with

○ Built a Scrapy web crawler (Python) with modular plugins for domain-specific data
ingestion, improving AI answer accuracy by 40% across 50+ client domains.
○ Led end-to-end development of a scalable AI/RAG system (the Knowledge Builder)
enabling enterprise clients to query knowledge from petabyte-scale datasets.
○ Designed and implemented L1/L2 processing pipelines to ingest unstructured data from
diverse sources (Dropbox, Google Drive, email, MCP, and web crawlers), generate vector
embeddings, and transform content into structured JSON for advanced analytics.
○ Built support for vector search, hybrid search, and context augmentation, enabling
interactive chat-with-your-data workflows for clients.
○ Developed and maintained distributed, asynchronous document processing pipelines
capable of handling petabyte-scale ingestion in parallel, ensuring high throughput and
reliability.
○ Led technical decisions and introduced best practices (TDD, documentation standards,
CI/CD), improving code quality and team productivity.
○ Tech stack: FastAPI (Python), Node.js, React, Milvus, MongoDB, Redis, Elasticsearch.

○ Managed patient scheduling, data entry, and coordination workflows using internal digital
systems and Excel-based tools.
○ Collaborated with cross-functional teams (nurses, drivers, management) to streamline
daily operations for 100+ patient appointments weekly.
○ Improved scheduling accuracy and reduced delays by organizing data-driven task tracking
and communication workflows.
○ Supported the digital transition of appointment and reporting systems, gaining exposure
to healthcare software operations and process automation concepts.

• Developed and delivered the HR/payroll module of a point-of-sale system for businesses, enabling staff to clock in/out and
track breaks via QR codes, with automated payroll calculation and reporting.
• Contributed to both the merchant-facing POS and an internal task management tool (built in-house to replace Monday.com),
improving operational efficiency and reducing software costs.
• Designed architecture, implemented features end-to-end, tested functionality, and collaborated directly with stakeholders
to align requirements and delivery.
• Tech stack: PHP, JavaScript, jQuery.

○ Developed and delivered the HR/payroll module of a point-of-sale system for businesses,
enabling staff to clock in/out and track breaks via QR codes, with automated payroll
calculation and reporting.
○ Contributed to both the merchant-facing POS and an internal task management tool
(built in-house to replace Monday.com), improving operational efficiency and reducing
software costs.
○ Designed architecture, implemented features end-to-end, tested functionality, and collab
orated with stakeholders to align requirements and delivery.
○ Tech stack: PHP, JavaScript, jQuery.

• Managed patient scheduling, data entry, and coordination workflows using internal digital systems and Excel-based tools.
• Collaborated with cross-functional teams (nurses, drivers, management) to streamline daily operations for 100+ patient
appointments weekly.
• Improved scheduling accuracy and reduced delays by organizing data-driven task tracking and communication workflows.
• Supported the digital transition of appointment and reporting systems, gaining exposure to healthcare software operations
and process automation concepts.