Pratham Soneja
ML EngineerAI is analyzing your overall score…
Identifying your key strengths…
Evaluating your skill match against the job requirements…
Assessing your cultural and operational fit
AI is analyzing your overall score…
Identifying your key strengths…
Evaluating your skill match against the job requirements…
Assessing your cultural and operational fit
AI Engineer · Agentic AI orchestration & Azure AI Systems · Edge Deployment & Computer Vision · Building AI that ships to production
I build AI systems that are measured by outcomes, not benchmarks. Currently at HCLTech — GPT API integrations and Azure-based AI infrastructure are part of it, but so is what's underneath: model foundations, agentic system design, orchestration pipelines, and the evaluation frameworks that tell you whether any of it actually works. 10,000+ active users in production. Before that: a computer vision pipeline on Jetson hardware, connected to a physical brake actuator, with a 150ms detection-to-actuation hard constraint. No cloud in the loop. No second chances on a bad inference. I work at the intersection of model engineering and business impact. The most expensive AI project I've been close to had 97% validation accuracy. Nobody used it — because the problem framing was wrong before the first line of code was written. I've also built systems where a 95% accurate model drove measurable cost reduction, because the error cost was mapped before the architecture was chosen. The difference between those two outcomes is not technical skill. It's knowing what question to answer before you build. That's what I bring to a project — whether it's a 3-month contract engagement or a full-time role building something that matters. Patent holder. Published. 3 years building AI under real constraints.
Manipal University Jaipur
Bachelor of Technology - BTech, Computer and Communication Engineering
August 1, 2020 – June 1, 2024
Manav Sthali School - India
senior school graduate, Computer Science
January 1, 2007 – January 1, 2020
HCLTech
Artificial Intelligence Engineer
October 1, 2024 – Present
Mahindra Group
Machine Learning Intern
September 1, 2023 – July 1, 2024
GeeksforGeeks
Data Science Intern
May 1, 2023 – June 1, 2023
Noida, Uttar Pradesh, India · On-site
Smollan
Data Science Intern
February 1, 2023 – April 1, 2023
Remote
AccioJob
Technical Content Writer
March 1, 2022 – July 1, 2022
Remote
SoftDodge
Data Analyst
January 1, 2022 – April 1, 2022
Jaipur, Rajasthan, India
PlotMyData
Data Scientest
September 1, 2021 – November 1, 2021
Remote
Varchasva
Community Contributor
August 1, 2021 – February 1, 2023
Jaipur, Rajasthan, India
Customer Churn Prediction
November 1, 2022 – November 1, 2022
Customer churn, also known as customer turnover or defection, refers to the loss of customers from a business. It is a major concern for companies, as it can have a significant impact on revenue and profitability. Therefore, predicting and preventing churn is an important task for businesses, and various methods have been developed to analyze customer data and identify patterns that may indicate an increased risk of churn. In this project, I focused on analyzing customer data to predict churn rates for a particular company. The data set included information about the customers' demographics, billing information, and support services provided by the company. First, I performed a statistical analysis of the data to identify any trends or patterns that could be related to churn. I used various visualization tools, such as histograms and scatter plots, to explore the data and gain a better understanding of the relationships between different variables. Next, I applied machine learning techniques to build a model that could predict the likelihood of a customer churning. I used logistic regression, a type of supervised learning method, to build the model. Logistic regression is a useful tool for predicting binary outcomes, such as whether a customer will churn or not. I used a portion of the data set to train the model, and the remaining data to test the model's performance. I also implemented hyperparameter tuning, which involves adjusting the model's parameters to optimize its performance. Through this process, I was able to achieve an accuracy of 81% in predicting the probability of a customer churning. The results of this project can be used by the company to identify key factors that may impact customer retention and to develop strategies to reduce churn rates.
US superstore sales analysis
July 1, 2022 – July 1, 2022
In this project, I analyzed the sales and profit data of a superstore in the United States from 2013 to 2016, and used Tableau to create interactive dashboards visualizing the results. The data set included information about the store's sales, profits, and various other metrics, such as the number of customers, the average sales per customer, and the average profit per order. To begin, I imported the data into Tableau and explored the different fields and their relationships. I used various visualization techniques, such as bar charts, line graphs, and scatter plots, to gain a better understanding of the data and identify any trends or patterns. Next, I focused on analyzing the sales and profit data to understand how they had changed over time and how they were affected by various factors. For example, I looked at the sales and profits by year, by month, and by region, and I analyzed the impact of different product categories and sales channels on the store's performance. Using Tableau, I created interactive dashboards that allowed me to easily visualize and compare the different data points. For example, I created a dashboard that showed the sales and profits by year and by region, and I added filters that allowed me to compare different regions or product categories. I also created charts and maps that visualized the sales and profits by state, which helped me to identify the store's strongest and weakest markets. Finally, I used the data and insights from this analysis to forecast the store's future sales and profits for the years 2017 and 2018.
Virtual Experiecne Program Participant
Accenture
June 25, 2026 – Present
Data Science
1Stop.ai
June 25, 2026 – Present
Machine Learning
Stanford Online
June 25, 2026 – Present
Machine Learning with Python
freeCodeCamp
June 25, 2026 – Present
Data science 2022: Complete Data Science and Machine Learning
Udemy
June 25, 2026 – Present
Python
Kaggle
June 25, 2026 – Present
Internship Program in Data Science
PlotMyData
June 25, 2026 – Present
Data Analysis with Python
freeCodeCamp
June 25, 2026 – Present
Intro to Machine Learning
Kaggle
June 25, 2026 – Present
CCNA: Switching, Routing, and Wireless Essentials
Cisco
June 25, 2026 – Present
CCNA: Introduction to Networks
Cisco
June 25, 2026 – Present
Intermediate Machine Learning
Kaggle
June 25, 2026 – Present
Cultural Fit Analysis
The candidate shows a strong interest in the ML/Data Science domain through diverse internships and personal projects. The variety of roles, from Data Analyst to AI Engineer (future role), and the range of project types (churn prediction, sales analysis, recommendation systems, disease prediction, fake news classification, Q&A bot) suggest adaptability and a broad interest in applying ML. However, the experience level (5 years) seems to be a mismatch with the actual professional experience (mostly internships and a future full-time role starting after graduation), which might indicate a junior profile rather than a senior one, potentially impacting cultural fit for a senior ML Engineer role.
Soft Skills & Operational Fit
The candidate's project descriptions indicate an ability to structure and explain technical work, suggesting reasonable communication skills. The 'Community Contributor' role at Varchasva implies some public speaking and collaboration experience. However, without psychometric test results, a comprehensive assessment of logical reasoning, work attitude, stress handling, and team collaboration is not possible.