SHARAD TIWARI

I'm

About

Machine Learning Engineer with 4 years of experience delivering production-grade ML and GenAI solutions across aviation, finance, and healthcare. Spearheaded a GenAI intent classifier that cut manual transcript review by 70% and built an MLOps pipeline reducing deployment time by 50%. Passionate about driving measurable business impact through scalable automation, real-time NLP services, and cloud-native architectures.

Strengths :
1. Highly organized and collaborative.
2. Initiative driven person delivering positive customer engagement.
3. Take extreme accountability and ownership of work.

Phone: +91 8269673369
Email: sharadktiwari2000@gmail.com

Interests

Data Science

Machine Learning

Natural Language Processing

Deep Learning

Computer Vision

Aviation Data

Speech Analysis

Forecasting

Education

Jiwaji University

B.Tech. in Computer Science And Engineering

July 2017 - May 2021

Relevant Coursework

Calculas, Linear Algebra, Statistics
Object Oriented Programming Systems
Data Structures & Algorithms
Database Management Systems
Operating Systems
Software Engineering
Theory Of Computation
Computer Architecture & Microprocessors

Data Techs

Diploma In Data Science

October 2019 - Janurary 2020

Relevant Coursework

Statistics
R Language
Python Programming Language
Machine Learning
Deep Learning
Natural Language Processing
Computer Vision
Data Mining

Online Certification

AWS Associate Machine Learning Engineer

AWS Cloud and Technical Essentials

Data Science on Microsoft Azure

Machine Learning Specialization

Deep Learning Specialization

Natural Language Processing in Microsoft Azure

A to Z Machine Learning with Python and R

Experience

Jet2 Travel Technologies

(Jet2.com & Jet2Holidays)

July 2024 - Present

Machine Learning Engineer

Developed a GenAI customer intent classifier (AWS Transcribe, Mistral fine-tuning, Snowflake Cortex AI) that reduced manual transcript review effort by 70%.
Forecasted aircraft revenue and bookings using XGBoost model (with r2_score 0.87 on validation data), optimizing route planning and cutting overbooking incidents by 30%.
Implemented Azure DevOps CI/CD pipelines with AWS ECR and Terraform to automate infrastructure provisioning and project deployments, reducing deployment time by 50%.
Built an end-to-end MLOps pipeline (AWS Glue, SageMaker, Lambda, Step Functions) and a hybrid feature framework (Snowflake, DBT) to automate model training and input standardization; piloted for 10+ ML projects.
Automated & cost effective document lookup with a “PDF Finder” pipeline, reducing work order processing time from 5–10 hours to 5 minutes.
Enforced data science best practices (checklists, code reviews, performance tuning & Azure DevOps artifact registry with Snyk Advisor security scans), reducing production bugs, accelerating model inference and vulnerability-free deployments.

Codiant - A Yash Technologies Company

January 2022 - Present

Machine Learning Engineer / Data Scientist

Collaborated with cross functional teams & Deployed a RAG-based Q&A pipeline (Azure OpenAI GPT-3.5, Ada embeddings, FAISS) that cut document lookup time from over 5 days to minutes.
Built an end-to-end document extraction system for aircraft maintenance ARC (EASA/FAA) documents using Azure Document Intelligence and a custom YOLO + OCR + Image Classification pipeline, achieving over 80% accuracy and enabling validation via a stakeholder-facing Streamlit app.
Designed a SpaCy NER + Elasticsearch + ML Classification pipeline to extract key details from aircraft maintenance records, achieving 93% field-level accuracy and automating record tagging (with f1_score 0.89) and labelling (with f1_score 0.93).
Engineered a resume parser (SpaCy NER & REL) deployed on AWS Lambda, auto-processing 1,000+ resumes/month across mid-sized firms and eliminating manual review and scoring effort for recruiters.
Developed a multimodal recommendation system using transformer-based embeddings (text, video, audio) for scalable cosine-similarity retrieval.
Deployed RASA-based chatbot platforms in healthcare and insurance domains, reducing customer response times by 60%.

Pojo Software Pvt Ltd

September 2021 - Janurary 2022

Quantitative Developer | Data Analyst

Designed and backtested high-frequency trading algorithms using statistical analysis and mathematical modeling, optimizing performance and reducing volatility.
Conducted data-driven market analysis to identify trading opportunities, driving improved portfolio returns.