Scott William

I am a Senior Full-Stack Engineer with over 8 years of experience delivering secure, AI-driven platforms across industries such as healthcare, fintech, insurance, and sports analytics. Specialised in the integration of large language models (LLM), retrieval-augmented generation (RAG), and distributed microservices on AWS and GCP. I have a proven record of shipping HIPAA-grade systems, event-driven Kafka pipelines, and high-traffic Next.js/React surfaces.

Experienced in architecting solutions from model serving on Kubernetes to typed TypeScript front-ends. I partner closely with product, data science, and security teams to deliver AI features that customers can trust in production environments.

Senior Full-Stack Engineer at DeepInspire (2023-06 – Present)

• Architected and delivered an AI-powered semantic search and recommendation system that integrated Python, Go, and TypeScript microservices, serving low-latency LLM-augmented results to production users.
• Integrated LangChain-based NLP and embedding pipelines for semantic and vector search across Pinecone, FAISS, and Weaviate, improving top-3 result precision by 34% and reducing median query latency to below 180 ms.
• Designed FastAPI and NestJS service layers for cross-service communication, ensuring typed contracts, retry mechanisms, and built-in observability for efficient system interactions.
• Provisioned multi-cloud infrastructure on AWS and GCP using Terraform, establishing locked-down IAM policies and reproducible environments for development, staging, and production.
• Containerized model-serving systems with Docker and Kubernetes, maintaining SLA performance even during 5x spike traffic during model launch events.
• Collaborated with data scientists and security teams to enforce data privacy and compliance (PII redaction, audit trails, and secure access controls) in model pipelines.

Senior Software Engineer at Timspark (2021-08 – 2023-05)

• Delivered secure full-stack healthcare applications integrating AI models for document processing, ICD-coding assistance, and anomaly detection across patient records.
• Designed Python/Flask and Node.js backend APIs that implemented event-driven, audit-grade data workflows with end-to-end tracing across the request lifecycle, ensuring seamless monitoring and reporting.
• Enhanced security across services by integrating fine-grained access control, data encryption, and continuous vulnerability scanning within CI/CD pipelines, strengthening the platform's compliance posture.
• Partnered with AI and product teams to integrate predictive analytics models into patient management systems, significantly reducing clinician documentation overhead and improving operational efficiency.
• Established AI-feature review processes (model validation, drift monitoring, and rollout gates), ensuring that features met the highest standards before reaching production.

Software Engineer at Smart Web Agency (2018-06 – 2021-07)

• Built secure insurance underwriting and patient-data workflows APIs in Python, .NET, and PostgreSQL, adhering to strict data validation standards and ensuring high-quality data management.
• Developed Kafka-based event-driven microservices that processed tens of millions of events per day, improving real-time transaction monitoring and ensuring predictable p99 latency for critical services.
• Integrated machine learning-driven anomaly detection models into backend services, reducing false-positive escalations to the analyst team by approximately 40%.
• Deployed reproducible infrastructure as code with Terraform and automated deployments using CI/CD pipelines, reducing release frequency from weekly to multiple times per day.
• Mentored junior engineers, facilitating onboarding for event-driven architecture and enforcing best practices for secure API design.

Bachelor of Science in Computer Science – University of Birmingham (2014 – 2018)