Hi, I'm Akshay Raju Kulkarni. Production ML for commercial impact & physics-driven AI for engineering.

AI Platform Engineer (Senior AI/ML Engineer)

Acting as founding-level AI engineer for Jumbo's company-wide AI platform to enable widespread AI adoption across business units, with full ownership over what gets built and why.

• Identified and prioritised the highest-leverage AI use cases across the business; currently spearheading two: AI-driven replenishment optimisation and assortment quality, targeting significant cost reduction, margin improvement, and operational efficiency.
• Defining the commercial case for each initiative, building the end-to-end systems, and owning the business outcome.
• Building memory and fact-retrieval systems for the Jumbo shopping assistant, enabling the AI to retain context and serve accurate, personalised recommendations across sessions.
• Integrating LLMs and agent frameworks into operational workflows while advising AI adopters and business teams on real-world limitations, benefits, and risks so adoption is grounded and sustainable rather than experimental.
• Working directly with leadership on AI strategy and platform direction, acting as the technical bridge between research, product, and cross-functional business teams.

Data Scientist / ML Engineer

Identified, built, and shipped end-to-end production ML systems across e-commerce and fulfilment that contributed >€2.5M in annual revenue uplift across the portfolio.

• Built and shipped production ML models for three high-value commercial problems: delivery slot availability, dynamic pricing, and regional capacity planning.
• Owned the full lifecycle from commercial problem framing and data pipeline engineering to feature engineering, model training (PyTorch, Scikit-learn), evaluation, deployment on AWS Lambda and Azure Databricks, and ongoing iteration.
• Applied computational optimisation and simulation thinking from aerospace to last-mile logistics scheduling under uncertainty.
• Operated with significant autonomy inside a large organisation, consistently initiating high-impact work rather than waiting for briefs.

Postdoctoral Researcher

Led research on AI-accelerated engineering design, sub-scale aircraft design, and aerodynamic surrogate modelling while supervising MSc researchers.

• Extended PhD research on sub-scale aircraft design and aerodynamic surrogate modelling—designing and running simulation campaigns to curate high-fidelity datasets for model training and supervising MSc students.
• Reduced aircraft engineering cycle times by 90% through automated simulation pipelines and ML-augmented design loops; evaluated AI physics models against strict engineering validation standards.
• Collaborated with Research In Flight (RIF) on FlightStream-driven surface-vorticity aerodynamics, contributing to peer-reviewed papers presented at international aerospace conferences.
• Applied Model Based Systems Engineering (MBSE) to formalise Knowledge Based Engineering (KBE) workflows for complex multi-disciplinary aerospace design — published in Engineering Informatics (with Bansal, La Rocca et al.), providing the structured system modelling underpinning digital twins and multi-physics simulation pipelines.

PhD Researcher

PhD in computational optimisation for sub-scale aircraft model design using physics-based simulation (Clean Sky 2, EU-funded).

• Built surrogate models over 5,000 to 10,000 physics simulations across 40 to 60 design variables to reduce expensive simulation costs and cut aircraft design lead times by 95% through surrogate-model-driven design space exploration.
• Core contribution: ML models that learn physical behaviour from simulation data and generalise to new aircraft designs — validated against wind-tunnel tests and real flight test data.
• Developed aerodynamic analysis methods combining panel methods, vortex-lattice solvers, and viscous-inviscid coupling to model wing performance and flight dynamics.
• Taught graduate courses in aerospace design, co-authored peer-reviewed publications in aerodynamics and applied simulation, and presented research at international AIAA and CEAS venues.

Similitude Augmentation in Sub-scale Flight Test Model Design. Computational optimisation, surrogate modelling, MDAO. 186 citations. Taught graduate courses in aerospace design.

Specialised in flight performance, propulsion systems, aerodynamic analysis, and aircraft design.

Core engineering foundation in structural mechanics, fluid mechanics, thermodynamics, and computational mathematics.