SAURABH RAMTEKE

DATA SCIENTIST | MACHINE LEARNING ENGINEER

SUMMARY
Machine Learning Engineer with 3+ years of experience developing and deploying data-intensive ML solutions for customer intelligence,
data activation, and business automation through distributed systems architecture. Driven by a growth mindset and motivated to solve
complex machine learning infrastructure challenges through innovative AI solutions. Specialised in Large Language Models (LLMs),
automated experimentation frameworks, predictive modeling, and performance marketing optimisation through cross-functional
collaboration. Approach problems with first-principles mindset while maintaining compassion for user needs and customer-centric
problem solving. Currently pursuing Executive Education in Product Management with Generative & Agentic.
PROFESSIONAL WORK EXPERIENCE
D ATA S C I E N T I S T, T U R I N G
REMOTE, INDIA
J U N E 2 02 4 - P RE SEN T
• Deployed AI Slack bot for automated knowledge retrieval and content generation, reducing manual search by 40% through intelligent
data collection and distributed systems architecture, demonstrating motivated approach to solving complex customer research
challenges
• Automated knowledge extraction from conversational data using router/specialist agent frameworks with machine learning
infrastructure, integrating Slack API for real-time, context-aware Q&A supporting customer data platform operations and data
activation for business operations.
• Employed SOTA LLM fine-tuning with supervised instruction training and adapter layers for performance marketing applications,
achieving 5% MMLU accuracy improvement through systematic automated experimentation frameworks and goal-oriented
optimisation.
• Aligned cross-functional teams and pitched innovative POCs, showcasing strong communication and analytical skills.
• Demonstrated agility by quickly adapting to new requirements, embracing feedback, and driving collaborative results in fast-paced
environments.
SO F T WARE EN G IN E ER , A R C ON
M U MB A I , I N DI A
J U N E 2 02 3 - J U NE 20 24
• Developed a hybrid model combining machine learning and heuristic methods to deliver actionable predictions and rule-based
insights For Customer Segmentation and Predictive Analytics.
• Led business understanding, user persona creation, and data validation, followed by comprehensive EDA, visualisation, and feature
selection—enabling informed decision-making and robust dashboard reporting reducing analytics turnaround time by 25%.
• Engineered scalable data handling pipelines utilising pagination, indexing, and Git versioning to manage large datasets and ensure
data integrity.
• Automated data collection workflows with APIs, multithreading, and SQLite, increasing data availability and reducing manual work by
10%, which resulted in measurable time and cost savings.
• Deployed interactive dashboards to track key business metrics, visualise model outputs, and quantify business impact, leading to a
17% improvement in operational efficiency and higher stakeholder satisfaction.
GRADUATE MACHINE LEARNING ENGINEER, CDAC
P U N E , I ND I A
M AY 2 02 2 - M AY 2 02 3
THESIS LINK
• Low-resource speech is first transliterated into high-resource speech. The resulting speech signal is then preprocessed through a
series of steps, including denoising, trimming, re-sampling, padding, and normalisation. The preprocessed signal undergoes feature
engineering using a Mel-spectrogram representation followed by spectral augmentation to enhance robustness. Relative positional
embeddings are added to the input feature map to incorporate positional information of speech segments. Convolutional Neural
Networks (CNNs) are employed for feature extraction, and the extracted features are subsequently passed through a Conformer block
integrated with an Adapter module. The model is optimised by minimising the Connectionist Temporal Classification (CTC) loss.
• Demonstrated an 18% reduction in word error rate (WER) in experimental settings, contributing to improved transcription accuracy
and forming the foundation for enhanced user experience in multilingual speech applications.

EDUCATION
E XEC U TIV E EDUCATI O N I N P RO DUCT M ANAG E M EN T WI T H G E N E R AT I V E & A G E N T I C A I
B IT S S C HO OL O F M AN AGEM EN T, M UM BAI, AUGUST 20 25 - PR ES EN T
M ASTER OF T ECHNO LO G Y I N DATA SCI ENCE
D E F E N C E I N S T I T U T E O F A D VA N C E D T E C H N O L O G Y, P U N E A U G U S T 2 0 2 1 - A P R I L 2 0 2 3
BACHELOR OF MECHANICAL ENGINEERING
S M T. K A S HI B A I N AVALE CO LL EGE OF ENGINEERING, P UN E A UGU S T 2014 - A U GU ST 2018

TECHNICAL SKILLS
Machine Learning & AI Infrastructure: Business & Strategy:
Machine Learning, Deep Learning, Large Language Models Product Management, Customer Research, Problem Definition,
(LLMs), Distributed Systems Architecture, Natural Language Cross-functional Collaboration, Growth Mindset, First-principles
Processing (NLP), Predictive Modeling, Statistical Analysis, AI Thinking, Agile Methodology, Performance Marketing.
Agents, RAG, Data Analysis, Data Visualisation,
Engineering: CERTIFICATIONS
Python, SQL, AWS, Apache Spark, Docker, REST API, - Machine Learning Foundations for Product Managers
LangChain, LlamaIndex, Google Agentic Framework, CrewAI, - Managing Machine Learning Projects
PyTorch, Pandas, Numpy, Scikit-Learn, SpaCy, Customer Data - Neural Networks and Deep Learning
Platform (CDP), Data Activation, Data Collection Pipelines, ETL/ - Artificial Intelligence
ELT, Performance Marketing Analytics, Budget Optimisation, A/B - Excel Skills for Business: Essentials
Testing, Web Scarping, Playwright. - Critical Thinking Skills for the Professional
- Apache Spark (TM) SQL for Data Analysts