Keshika Sharma

Business Development Executive

Pune, Maharashtra, India2 yrs 3 mos experience

Key Highlights

Expert in data-driven storytelling and insights.
Developed a predictive system for neonatal jaundice.
Skilled in machine learning and data visualization.

Stackforce AI infers this person is a Data Scientist with a focus on Healthcare and Machine Learning.

Contact

Skills

Core Skills

Data ScienceMachine Learning

Other Skills

Analytical SkillsAnalyticsBig DataC (Programming Language)C++CommunicationComputer VisionContent CurationData AggregationData AnalysisData AnalyticsData CleaningData CollectionData EthicsData Integrity

About

I’m a data-driven storyteller, insight enthusiast, and business problem-solver. I specialize in uncovering patterns within data to drive smarter decisions and actionable strategies. For me, data isn’t just numbers—it’s a powerful tool to understand customer behavior, optimize operations, and uncover opportunities that fuel business growth.

Experience

2 yrs 3 mos

Total Experience

10 mos

Average Tenure

Current Experience

Aristiec

Business Analyst

Jun 2025 – Jul 2025 · 1 mo · Gurugram, Haryana, India · Remote

Tata communications

Python Developer

May 2025 – Jul 2025 · 2 mos · Pune, Maharashtra, India · On-site

Prometeo iit jodhpur

2 roles

Head of Public Relations

Oct 2024 – Jan 2025 · 3 mos · Jodhpur, Rajasthan, India

Public Relaions Volunteer

Nov 2022 – Oct 2024 · 1 yr 11 mos · Jodhpur, Rajasthan, India

Tedx iitjodhpur

2 roles

Head Curator

Aug 2024 – Oct 2024 · 2 mos · Jodhpur, Rajasthan, India

Assistant Head Curation

Nov 2023 – Aug 2024 · 9 mos · Jodhpur, Rajasthan, India

Indian institute of technology jodhpur

Undergraduate Student Researcher

Aug 2024 – Feb 2025 · 6 mos · Jodhpur, Rajasthan, India

Situation: Tackled the challenge of neonatal jaundice in low-resource settings by exploring a non-invasive, cost-effective diagnostic approach.
Task: Develop a sensor-integrated machine learning system to accurately predict bilirubin levels in newborns.
Action:
Simulated the system circuit in LTspice using actual component specifications to ensure design feasibility.
Calibrated the MAX30102 optical sensor, achieving reliable SpO2 readings across diverse environments.
Built real-time visualization dashboards using Matplotlib and Seaborn for sensor data analysis.
Trained and fine-tuned an SVR model using GridSearchCV for precise bilirubin prediction.
Result:
Achieved 98.55% SpO2 accuracy with just 6.52 variance.
Improved medical data interpretability by 30% through visual dashboards.
Final SVR model achieved an R² score of 0.784, demonstrating strong clinical potential.