Mrinal Kamboj

Supply Chain Optimization and Forecasting

Products: Inventory, price, promotion forecasting, and planning systems.

• System design, architecture, technical roadmap, performance, concurrency, and scalability.

Clients:

• Ralph Lauren, Dollar General, Carters, Cracker Barrel, Bealls, and Kik, with millions of SKUs allocated, have cumulative savings of over $500 million.

AI Impact:

• Leveraging AI tools for higher-quality code and architecture with guardrails. Building AI memory context for faster shipping.
• Building AI agents to automate allocation and pricing workflows.

Team:

• Manage 10 architects and EMs (100+ indirect), including database, DevOps, and data ingestion. Built and scaled teams from scratch.

Technical Impact:

• Backend architecture, scalability, performance optimization, debugging, and frameworks.
• Scaled to update 1 billion+ records in BQ in less than 10 seconds.
• For Plansmart cube creation with millions of data points in less than 5 sec using Duck DB and Parquet.
• Database: Planning, Query Tuning, Replication, Partitioning.
• Microservices: Designing Bounded Contexts, Asynchronous Communication (Event and Orchestration-Based Patterns)
• DevSecOps - Infrastructure planning, deployment (Ansible), containers (Packer), Kubernetes, and automation. SAST, DAST, SCA, Penetration Testing, Cloudflare, Patch Management, Threat Modeling, and Auditing.
• Cloud FinOps: Extensive automation and planning consistently saved $4 million annually.
• System observability (APM, logging, and infrastructure monitoring) using DataDog and Grafana.

Revenue:

• Revenue growth from $10 million to $50 million.

Clients: SAIA, Holland, YRCF, FedEx, ABF, BHP.

Haul and Rail Projects:

• Architecture, system design, product roadmap, and client interaction.
• Framework design (Data access, Optimization solvers, Caching, Logging, Security) (Productivity Impact: 50%). Profile and resolve the performance and memory issues.
• Designed a framework using expression trees, parallelism, filtering, and sorting over 3 million records in under 500 ms (Performance Improvement: 95%, Client: FedEx, Revenue: $2 million). A custom object pool to integrate with Exasol IMDB (Client: Holland, Revenue: over $1 million).

• Designed an automated stress test framework for the Oracle data access stack via ODP.NET. High issue-fix rate of 75%.