Mohsin Sheikhani

Designed an event-driven multi-agent financial intelligence system with autonomous anomaly detection, natural language querying via LLM tool-use, and predictive cash flow forecasting, orchestrating agent-to-agent communication with human-in-the-loop approval workflows.
Developed automated evaluation pipelines using LLM-as-a-judge scoring, deterministic test harnesses.

Designed and shipped AI-driven backend services integrating OpenAI for text generation and image synthesis, automating enterprise workflows and reducing manual design tasks by about 40%. Built serverless microservices on AWS (Lambda, API Gateway, S3, RDS/PostgreSQL) and introduced patterns that cut infrastructure costs by 30%.

Delivered AI automation solutions and end-to-end Node.js microservices integrating LLMs and Amazon Polly, accelerating content delivery and improving throughput by 40% while cutting operational costs by 25%. Optimized PostgreSQL schemas and implemented Redis caching to lower average API response from 800ms to <200ms.

Architected and launched a SaaS AI analytics platform in under 4 months using TypeScript, Node.js, AWS Lambda, and RDS, enabling first customer adoption within 2 weeks and handling 1M+ monthly transactions with 99.99% uptime. Automated infra provisioning with AWS CDK and improved observability with CloudWatch, reducing MTTD by 40%.

Built customer-facing SaaS features with Next.js and Node.js/PostgreSQL, contributing to 40% YoY user growth and implementing OAuth2/JWT via Amazon Cognito to reduce unauthorized access attempts by 90%. Led migration of on-prem PostgreSQL to Amazon Aurora using AWS DMS, cutting infrastructure costs by 40% and operational burden by 70%.

Completed a Bachelor of Computer Science at DHA Suffa University in Karachi, graduating in January 2022.