I'm Aakshat Agarwal, a data and technology professional from Jaipur working with Python, SQL, machine learning, business intelligence, cloud tools and backend technologies.
Python · SQL · Machine learning · Five documented systems
Proof, not positioning
80.4% of churners caught, at roughly one real save per two calls
Working principle
Start with the decision. Trace the evidence. Build only what makes the result easier to use.
I enjoy converting raw information into useful products, insights and automated solutions. My work spans data analysis, machine learning, dashboard development, API creation and data-pipeline design. I am especially interested in projects where analytical thinking and software development work together to solve a practical problem.
Alongside technical development, I am strengthening my understanding of AI, machine learning, Python, SQL, data engineering and backend systems through continuous learning and project-based practice.
7,043 customers → churn probability
churners caught
Selected systems
Each chapter moves from an unreliable input to a decision someone can use—diagnosis, fire risk, retention, document matching, or a month of bills filed without anyone watching.
Classifying breast masses where a missed cancer costs more than a false alarm.
An unattended pipeline that collects, reads and files 34 utility bills a month.
Predicting Algerian fire days from weather alone, after removing the leakage.
Catching four in five churners, and being honest about the cost of that.
NLP app scoring resume–job fit with a skill-gap report.
A practical inventory of tools used to inspect data, build models, expose results and keep the delivery path reliable.
2 strong-use tools
3 strong-use tools
1 strong-use tools
2 strong-use tools
Developing range
3 strong-use tools
Signal path
Reliability comes before modelling. The path stays visible from the first row of raw data to the final prediction, dashboard or API response.
Raw Data
verified stage
Cleaning
verified stage
Exploration
verified stage
Feature Engineering
verified stage
Model / Analysis
verified stage
API / Dashboard
verified stage
Insight / Prediction
usable output
Experience
Employment, kept separate from the project narratives and the job simulations—this is the work with a colleague waiting on the output.
Kalash Group
Jaipur, Rajasthan, India
Formal computer-science study, strengthened through project work that moves from analysis into shipped interfaces and APIs.
University of Rajasthan
Jaipur, Rajasthan, India
University of Rajasthan
Jaipur, Rajasthan, India
Parallel learning track
Verified learning
Professional exercises and certifications are kept distinct from employment—the value is in the work performed, not a borrowed logo.
Amazon Web Services · via Forage
Completed a practical simulation covering cloud architecture, EC2, S3, RDS and cost-optimisation strategies.
Deloitte Australia · via Forage
Analysed business datasets and produced data-driven recommendations, including a modelled scenario of ≈20% operational-efficiency improvement. Completed as a job simulation, not employment at Deloitte.
JPMorgan Chase & Co. · via Forage
Completed a practical simulation focused on enterprise software-development workflows and engineering tasks. Completed as a job simulation, not employment at JPMorgan Chase.
Certification
Covered foundational data-science concepts, algorithms, statistical modelling and structured problem-solving.
be10x
Completed a workshop focused on practical AI tools, ChatGPT and AI-assisted presentation creation.
Repositories hold the implementation detail behind the project narratives—data preparation, modelling choices and delivery code.
@Aaksh26at
Churn classifier over 7,043 customers catching 80.4% of churners, with the false-alarm cost reported rather than hidden.
Live learning queue
Deepening production-ready ML workflows and model lifecycle.
Strengthening data structures and algorithmic fluency.
Query optimisation, warehousing and pipeline design.
API design, serialization and deployment patterns.
Open to data, AI/ML, Python, backend and analytics opportunities—and conversations where analytical thinking needs to become a useful product.