← Back to aboutResume

Machine Learning Engineer

Machine Learning Engineer with a focus on applied machine learning, deep learning, and model development, with experience across time-series forecasting, computer vision, NLP, and large language models. Experienced in developing, fine-tuning, evaluating, and deploying ML models using PyTorch, Hugging Face, and Scikit-learn, with 3+ years of production software engineering experience at Microsoft. Currently pursuing an MS in Artificial Intelligence at Nanyang Technological University (NTU), Singapore.

SingaporeTwissa Modi

Download PDF

01 / Experience

07/2026 – 12/2026

Software Engineer Intern – AI Agents

Robert Bosch (SEA) Pte Ltd
Singapore

Designing an NL-to-SQL analysis system via Copilot Studio custom connectors to on-prem services, grounding query generation in database metadata to enable natural-language cost-centre spend analysis for business users
Building an Agent-to-Agent (A2A) bridge between Copilot Studio and Gemini Enterprise agents, including secure Azure AD authentication, to enable cross-platform data and workflow interoperability

08/2022 – 07/2025

Software Engineer

Microsoft
Hyderabad, India

Led end-to-end telemetry implementation for Windows Share Sheet, defining product usage metrics and partnering with the Data Science team to deliver actionable analytics dashboards.
Re-architected backend services for Microsoft Family Safety using C#/.NET, reducing compute costs by 16%, and developed the App Category feature for more granular application controls.
Co-invented and filed a U.S. patent, “Drop Tray Engine(s) for Providing a User-Based Drag-and-Drop Action Experience” (Docket 503629-US01; filed Feb 2025), for a Windows feature reaching millions of users.

02/2022 – 07/2022

Software Engineer Intern

McKinsey Digital Labs
Bangalore, India

Built interactive data visualization dashboards using React.js for client-facing applications, translating complex datasets into actionable insights for business stakeholders.

02 / Education

08/2025 – 12/2026

Master of Science in Artificial Intelligence

Nanyang Technological University
Singapore

GPA: 4.7/5.0 Relevant Coursework: Machine Learning, Deep Learning, Natural Language Processing, Large Language Models, Computer Vision

07/2018 – 06/2022

Bachelor of Engineering in Computer Engineering

Thapar Institute of Engineering & Technology
India

CGPA: 9.54/10 Scholarship holder

03 / Projects

Day-Ahead Electricity Load Forecasting

Time Series Forecasting | Probabilistic Machine Learning

Built a day-ahead electricity load forecasting pipeline using ENTSO-E grid data, achieving 2.65% MAPE with calibrated probabilistic forecasts using Conformalized Quantile Regression (CQR) and 89.9% prediction interval coverage.
Deployed an automated forecasting pipeline with daily data ingestion, model inference, and monitoring using GitHub Actions and Supabase.

Rossmann Sales Forecasting

Time Series Forecasting | Machine Learning

Built an end-to-end LightGBM sales forecasting pipeline for the Rossmann Kaggle competition, using Optuna hyperparameter optimization, log-transformed targets, and leakage-safe temporal features to achieve 0.1199 RMSPE on the held-out test set.
Engineered lag and rolling features (49/56/63 days) alongside promotion and competition-period features to capture temporal, seasonal, and business-driven demand patterns.

Serengeti Wildlife Classifier

Deep Learning | Computer Vision

Fine-tuned ResNet18 for 11-class wildlife image classification using transfer learning, data augmentation, and class-balanced sampling, achieving 84.0% test accuracy on a held-out split grouped by camera site.
Applied GradCAM for model interpretability and deployed the inference service using FastAPI and Docker for reproducible serving.

Banking Intent Classifier

Efficient LLM Fine-tuning

Fine-tuned Qwen2.5-0.5B on BANKING77 using LoRA and supervised fine-tuning (SFT), achieving 84.6% accuracy while training only 0.12% of model parameters.

04 / Technical skills

Programming & Data

Deep Learning & Computer Vision

Infrastructure

Machine Learning

LLM & NLP

Agentic AI & RAG

05 / Certifications & activities

Machine Learning SpecialisationDeepLearning.AI (Coursera)

Foundations of AIMicrosoft

VolunteerWomen Devs SG

MemberWomen in Tech – Microsoft

I am most useful where a model needs to become a dependable part of a larger system.