Mayank Sharma

About Me

I'm a Data Scientist with 5+ years of experience building machine learning and analytics solutions across healthcare, manufacturing, and enterprise systems. I combine deep ML and NLP expertise with strong analytics capabilities to deliver intelligent systems that drive measurable business impact. My work spans Generative AI, LLMs, model optimization, operational analytics, and production ML systems.

M.Tech CSE - Data Science, IIT Jammu · India

My Story

My journey began with machine learning. During my M.Tech in Data Science at IIT Jammu, I developed a deep fascination for how models can understand language, extract insights, and solve real-world problems. This foundation shaped my approach: start with the problem, build the model, measure the impact, and iterate.

At nference.ai, I transitioned cutting-edge NLP research into production healthcare systems that clinicians trust. This taught me the critical difference between research accuracy and production reliability. I optimized models for real-world constraints (inference speed, model size, data scarcity), built evaluation frameworks that matter, and learned that rigorous evaluation is non-negotiable. Achievements: 9× inference speedup, 8× model compression, few-shot prompting reducing labeled data reliance by 99%.

At Lamipak, I built GenAI systems and analytics platforms serving thousands globally. I developed LamiOps (operational analytics + AI assistant for 10K+ users, 99% efficiency gains) and LamiTracker (regulatory intelligence across 200+ sources, 18+ countries, 95% automation). This solidified my belief: ML + Analytics together drive business value.

Today, I position myself as a Data Scientist with strong analytics capabilities. I combine machine learning expertise (NLP, GenAI, LLMs, prompt engineering, model optimization) with analytics (KPI definition, experimentation, insights). Beyond client work, I write Token by Token — technical deep-dives on ML, NLP, AI, and building systems that matter.

My Journey

From ML fundamentals to production research to analytics systems at scale — how I evolved from researcher to Data Scientist with analytics expertise.

Machine Learning Foundations

M.Tech in Data Science from IIT Jammu (CGPA: 8.70/10). Built ML models and explored statistical approaches to solving real problems. Mentored 100+ students in Algorithms, Software Tools, and Web Engineering.

Research to Production ML

At nference.ai, transitioned healthcare NLP from research into production systems clinicians trust. Optimized models (9× speedup, 8× compression), built evaluation frameworks, demonstrated few-shot prompting effectiveness (99% reduction in labeled data). Learned that production constraints and evaluation rigor drive decisions.

ML + Analytics at Scale

At Lamipak, built GenAI systems and analytics platforms. LamiOps (operational analytics + AI, 10K+ users, 99% efficiency), LamiTracker (regulatory intelligence, 200+ sources, 18+ countries, 95% automation). Combined ML expertise with analytics to drive measurable business outcomes.

Impact & Achievements

AAAI 2024 Publication

Co-authored OpenMedLM (AAAI 2024, NEJM AI), demonstrating that prompt engineering can match or exceed fine-tuning performance in medical question answering. Validates practical ML approaches for building effective AI systems with open-source models.

Model Optimization Excellence

Optimized Transformer models (BERT, BioClinicalBERT) achieving 9× inference speedup and 8× compression while maintaining accuracy. Received Bravo Award at nference for making real-time ML deployment feasible in healthcare production systems.

Few-Shot Learning Strategy

Demonstrated few-shot prompting could replace supervised annotation pipelines, reducing labeled data reliance by 99% while maintaining extraction accuracy. Key innovation that shaped data strategy for scaling ML systems across diverse client distributions without manual labeling.

GenAI Systems at Scale

Built LamiOps (operational analytics & AI assistant serving 10,000+ daily users), achieving 99% reduction in manual troubleshooting time. Deployed production GenAI system handling multilingual, multimodal retrieval at manufacturing scale.

Global Analytics Intelligence

Engineered LamiTracker, aggregating regulatory and competitive intelligence from 200+ sources across 18+ countries and 9+ languages. Achieved 95% automation of manual research workload through intelligent extraction and temporal analysis.

Technical Thought Leadership

Teaching Assistant at IIT Jammu (2019-2021) mentoring 100+ students. Author of Token by Token, publishing technical deep-dives on ML, NLP, AI, and the practical engineering behind production systems.

Beyond the Code

Built on hard work, powered by a touch of talent, sprinkled with humor, and fueled by infinite nerdiness.

Staying Active

When I'm not debugging models, you'll find me balancing code with cardio — gym sessions for strength, running for endurance, and cycling for exploring new trails. Movement keeps the mind sharp and the ideas flowing.

Fledgling Bookworm

A growing passion for reading keeps me curious beyond technical papers. Whether it's exploring new ideas, learning from diverse perspectives, or simply unwinding with a good story, books are becoming an essential part of my routine.

Grounded in Mindfulness

Meditation helps me stay centered amidst the fast-paced world of AI. It's not just about relaxation, it's about clarity, focus, and maintaining perspective when tackling complex problems.

One thing to remember about me: my humility isn't flattery — it's a value shaped by a humble upbringing.

Skills & Expertise

Machine Learning & AI

Generative AI & Large Language Models Natural Language Processing (NLP) Deep Learning & Neural Networks Prompt Engineering & Optimization Fine-tuning & Model Adaptation (LoRA, PEFT) Model Optimization (Compression, Quantization) PyTorch & Transformers Statistical Modeling & Analysis

Analytics & Business Impact

KPI Definition & Tracking Business Analytics & Insights Operational Analytics at Scale Experimentation & A/B Testing Evaluation Frameworks & Rigor Power BI & Dashboarding Data-Driven Decision Making Metrics & Impact Measurement

Supporting Technologies

SQL & Data Analysis Vector Databases (FAISS, Qdrant) Multimodal RAG Systems AWS & Docker Deployment CI/CD & ML Deployment Data Modeling (supporting) PostgreSQL, MongoDB Python & Software Engineering