Deepak Sai Pendyala
Data AnalystAI 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 Scientist Intern @Pearson | MS. CS @NC State | Applied Science @Amazon
CS guy passionate about applying AI and Software Engineering to solve real-world problems, currently pursuing a Master’s in Computer Science at North Carolina State University. As a Graduate Research Assistant at the STEPS Center, I develop scalable, explainable multi-agent LLM systems and Retrieval-Augmented Generation (RAG) architectures with trust, validation, and real-time feedback. Previously, I worked on a DARPA-funded project at CAMAL Labs, building real-time machine learning and computer vision pipelines for additive manufacturing. My work reduced inference time by 4.3x and improved model accuracy and efficiency by over 70%, supporting first-time-right production of defense-critical components. I also gained experience as an Applied Scientist Intern at Amazon, developing ML and GenAI solutions. With a Bachelor’s in Electrical and Computer Engineering from India, I founded and led the Intel IoT Club, mentoring 35+ members in a 2000+ member community. I am open to Full-time opportunities (May 2026) in Data Science, Applied Science, Machine Learning, and AI, especially in areas intersecting with Computer Networks, IoT, and Software Development. Let’s connect if you're looking for someone eager to make an impact in these domains.When I am not coding or tinkering with devices, I enjoy cooking, photography, blogging, and music. You can find some of my work on my GitHub profile.
North Carolina State University
Master of Science - MS, Computer Science
August 1, 2024 – May 1, 2026
Amrita Vishwa Vidyapeetham
Bachelor of Technology - BTech, Electrical and Computer
October 1, 2020 – May 1, 2024
Pearson
AI Scientist
June 1, 2026 – Present
Hoboken, NJ · Remote
SproutsAI
AI Engineer
June 1, 2025 – August 1, 2025
Milpitas, California, United States · Remote
STEPS (Science and Technologies for Phosphorus Sustainability) Center
Graduate Research Assistant (Applied AI)
May 1, 2025 – June 1, 2026
Raleigh, NC · On-site
NC State Center for Additive Manufacturing and Logistics (NCStateCAMAL)
Research Assistant (Machine Learning)
September 1, 2024 – May 1, 2025
Raleigh, North Carolina, United States · On-site
Amazon
Applied Scientist
July 1, 2023 – December 1, 2023
Gurugram, Haryana, India · On-site
ActiveRAG Next: Multi-Agent Reasoning System with LangGraph
May 1, 2025 – May 1, 2025
Built an end-to-end Retrieval-Augmented Generation system with agent-based coordination, reasoning, validation, and feedback in 3 days. Integrated Streamlit UI with LangGraph for real-time multi-agent execution, dynamic query routing, and interactive feedback loops. Used RAG-Fusion retrieval, knowledge graph extraction, and Graph-of-Thought reasoning for explainable outputs. Enabled stateful chat, execution trace, and validation-driven reruns with confidence scoring and user feedback routing.
Detecting and Mitigating Bias in Fraud Detection Models
October 1, 2024 – December 1, 2024
This project focuses on developing a hybrid fraud detection model to ensure fairness and accuracy in identifying fraudulent bank account applications. By addressing issues such as class imbalance and bias in datasets, the model achieves equitable outcomes across demographic groups while maintaining high predictive performance. https://github.com/deepaksaipendyala/Detecting-and-Mitigating-Bias-in-Fraud-Detection-Models-A-Fairness-Aware-Approach
Predictive Model-Based Power Price Tagging Under Deregulated Environment
October 1, 2023 – May 1, 2024
During this project, I developed an innovative auction mechanism leveraging genetic algorithms and machine learning models to optimize power generation resource allocation and accurately predict Market Clearing Prices (MCP) in a deregulated environment. Using data-driven approaches, I achieved high-accuracy power prediction and Economic Load Dispatch (ELD) optimization, utilizing comprehensive data from the Indian Energy Exchange (IEX). The project demonstrated superior performance metrics for LSTM models, providing actionable insights for efficient power price tagging and strategic bidding, thereby enhancing market competitiveness and profitability.
Smart Car Parking System using SCADA Network
June 1, 2023 – July 1, 2023
Developed a smart vehicle parking system using sensors to detect available slots, display the best slots to drivers, and control the system over a SCADA network. Features: Real-time monitoring and control of parking slots using a SCADA system. Central server (Laptop) with remote clients (Two Raspberry Pico) equipped with IR sensors. Utilized socket programming for SCADA master setup with multithreading. Data processing and communication between server and Raspberry Pico. Implemented gate control, LED displays, and data updates in the database and GUI. System Flow: Vehicle approaches the Pico at the entry gate.SCADA master communicates slot availability. Displays available slots at the entrance. Updates and communicates slot occupancy. If all slots are filled, ""NO SLOTS AVAILABLE"" is displayed.
Recommender systems in Online Reviews of Airlines - NLP
October 1, 2022 – October 1, 2023
Details: Conducted research on identifying key drivers of customer satisfaction in online reviews of airlines using a hierarchical topic modeling approach. Collected a large dataset of online reviews from various sources and preprocessed the data by removing noise such as stop words and punctuation. Tools/Area: Latent Dirichlet Allocation (LDA), sentiment analysis, machine learning (clustering, classification) Outcome: Successfully implemented LDA for topic modeling and identified key drivers of customer satisfaction in online airline reviews. Evaluated sentiment associated with each driver and provided insights into strengths and weaknesses of airlines in meeting customer needs and expectations. Made recommendations for airlines to enhance their services and address customer concerns.
Credit Score Classification
October 1, 2022 – January 1, 2023
Details: Developed a machine learning model for accurately classifying the credit score of individuals using K-NN, PCA, SVM, and Neural Networks. Applied K-NN algorithm for data classification, PCA for dimensionality reduction, and Neural Networks. Tools/Area: Python, K-NN, PCA, SVM, Neural Networks, Softmax Function, Binary Sigmoid Function, Cross-Entropy Function. Outcome: Successfully built an accurate and efficient model for credit score classification using a combination of machine learning techniques. Achieved reliable credit score classification results, providing financial institutions with valuable insights for making informed decisions on lending and creditworthiness. "
File Tagger
July 1, 2022 – July 1, 2022
Tag: - A tag is an arbitrary value that can be associated with a file. (Like names of people, places, events, etc.) - Users should be able to tag their files and locate them using tags. - A file can be associated with multiple tags, and a tag can be associated with multiple files. - You can relate it to something like how an email-client help organize emails using labels. Features of the software: - User can select a file or multiple files in their windows operating system, Add a tag to the file. - User can search the files using the tags assiociated with them. Tech Stack Used: - Python3 - MySQL Cross Platform Compatibility: - Implemented in Windows. - It can be developed for Other OS Like Linux, MacOS, etc. too. - Open Sourced Project, Pull requests are accepted.
Crop Prediction Machine Learning model
October 1, 2021 – October 1, 2021
- Developed a crop prediction machine learning model using LightGBM classifier and trained it using a dataset from Kaggle - Created a website using HTML, CSS and JavaScript for the front end, and python flask for the backend. - Hosted the machine learning model on the website Problem Statement: 50% of Indian population is dependent on agriculture for livelihood. One of the major factors that affect agriculture is the quality of soil. Soil supplies essential nutrients for the crop to grow and flourish. Choosing a crop that doesn't fit with the soil type not only degrades the quality and quantity of the crop , but also deteriorates the soil of its nutrients. Unhealthy crop patterns will also damage the quality of soil which in turn again affects the yield. Our Solution: To tackle this problem, our team has come up with a project that could help people check the soil condition and also gives crop recommendations depending on the soil parameters like nutrients in the soil, moisture in soil and rainfall. This project is based on a highly trained ML model to keep the recommendation as accurate as possible. The solution can be easily implemented and accessed by people. The user have to enter the soil parameters to check his soil compatibility and it is absolutely cost effective. Technical Stack Description: We obtained a dataset from Kaggle and then we took LightGBM classifying machine learning model and trained it using the dataset, later we created a website using HTML, CSS and JavaScript for front end and python flask for the backend, then we hosted our ML model on the website. The user should enter the soil parameters and a recommended crop will be displayed with given soil parameters on the website. User and Admin will get an email with same crop info only when name and email id is given.
Personal Portfolio Website Development
July 1, 2021 – July 1, 2021
Description: Developed a personal portfolio website. Successfully built the back-end using Python Flask. Currently enhancing the front-end leveraging basic HTML, CSS, and JS, along with the use of templates. The website features fun elements like facts, memes, and jokes. Features: Automated confirmation and greeting emails for form submissions. Receipt of form details to a personal email as a reminder. Form details are automatically saved to CSV and txt files. Provided an educational section explaining how the internet works and a beginner's guide to setting up a website. Skills: Web Development, Python Flask
Email Bot
June 1, 2021 – June 1, 2021
This Program will automate the process of sending emails to thousands of people through python. works with smtp - simple mail transfer protocol.
TwitterFollowbackBot
June 1, 2021 – June 1, 2021
Twitter Automation: Auto Followback bot, displays Your public tweets, analyse the tweets with keyword and n of top tweets as a result. works with Twitter API. Require twitter developer account
Password Breach
June 1, 2021 – June 1, 2021
This program checks whether your password leaked in any Data Breaches. Your can even do it at (https://haveibeenpwned.com/Passwords) but Your password travels over internet right? so, its not safest method. This works in most efficient way in your offline pc. When You enter a password, it will be hashed with sha1 first 5 letters of hash code will be sent to haveibeenpwned API and receives a Dict of password Data with those first 5 characters of hashed password then this program crosscheck with the received hash and finds whether your password leaked, if so how many times . else, it shows good to go.
sms bot
June 1, 2021 – June 1, 2021
This Program will automate the process of sending sms to thousands of people through python. We can Customize this in our own way. works with Sms Api.
File & Image Processing
June 1, 2021 – June 1, 2021
Image processing : JPG to PNG Converter using PYTHON3 (https://github.com/deepaksaipendyala/JPGtoPNG), simple image editor with pillow module in Python3: (https://github.com/deepaksaipendyala/image) File Processing: PDFmerger: (https://github.com/deepaksaipendyala/PDFmerger)
waste-management
May 1, 2021 – June 1, 2021
Waste management Program For municipality to Compute statistics of the waste collection per area,reward the house which gives less quantity of waste,total earning for the municipality per month,what percentage of the waste is from what category and As an individual family,the amount spending towards this. I wrote an Algorithm and successfully implemented it using C-Programming for the above-mentioned scenario.
WiFi Jammer/WiFi deauther
September 1, 2018 – September 1, 2018
This software allows you to easily perform a variety of actions to test 802.11 wireless networks by using an inexpensive ESP8266 WiFi SoC (System On A Chip). The main feature, the deauthentication attack, is used to disconnect devices from their WiFi network. No one seems to care about this huge vulnerability in the official 802.11 WiFi standard, so I took action and enabled everyone who has less than 10 USD to spare to recreate this project. I hope it raises more attention on the issue. In 2009 the WiFi Alliance actually fixed the problem (see 802.11w), but only a few companies implemented it into their devices and software. To effectively prevent a deauthentication attack, both client and access point must support the 802.11w standard with protected management frames (PMF). While most client devices seem to support it when the access point forces it, basically no WiFi access point has it enabled.
Complete Python Developer in 2021: Zero to Mastery
Udemy
June 25, 2026 – Present
Microsoft AI Classroom Series
Microsoft
June 25, 2026 – Present
AWS Machine Learning Foundations
Udacity
June 25, 2026 – Present
Intel® Student Ambassador Program IOT AMBASSADOR LEVEL 1
Intel
June 25, 2026 – Present
Operating Systems and You: Becoming a Power User
Coursera
June 25, 2026 – Present
System Administration and IT Infrastructure Services
Coursera
June 25, 2026 – Present
IT Security: Defense against the digital dark arts
Coursera
June 25, 2026 – Present
Google IT Support Specialization
Coursera
June 25, 2026 – Present
Intel® Student Ambassador Program IOT AMBASSADOR LEVEL 2
Intel
June 25, 2026 – Present
Build a Face Recognition Application using Python
HCL GUVI
June 25, 2026 – Present
Intel Edge AI Certification
Intel
June 25, 2026 – Present
MATLAB Onramp
MATLAB Coding
June 25, 2026 – Present
OpenVINO Training
Intel
June 25, 2026 – Present
Technical Support Fundamentals
Coursera
June 25, 2026 – Present
The Bits and Bytes of Computer Networking
Coursera
June 25, 2026 – Present
AI for India
HCL GUVI
June 25, 2026 – Present
30 Days of Google Cloud
June 25, 2026 – Present
Simulink Onramp
MATLAB Coding
June 25, 2026 – Present
Problem Solving (Basic) Certificate
HackerRank
June 25, 2026 – Present
Cultural Fit Analysis
The candidate's project portfolio is extensive and diverse, covering areas from web development and automation to advanced AI/ML applications. This breadth indicates a strong curiosity and a drive to learn and apply new technologies. While many projects align with data science and machine learning, the target role is 'Data Analyst'. Some projects, like 'WiFi Jammer' or basic web development, show a broad interest but might not directly contribute to a senior data analyst profile. The candidate's academic background and internships at companies like Amazon and Pearson suggest an ambition for high-impact roles. The numerous certifications also highlight a commitment to continuous learning. However, the sheer volume and variety of projects, some less relevant to a senior data analyst, could indicate a lack of focused specialization for this specific role.
Soft Skills & Operational Fit
The candidate's project descriptions indicate a proactive and problem-solving attitude, with a focus on building practical solutions. The diverse range of projects suggests adaptability and a willingness to explore different technical areas. However, without direct assessment data on soft skills, a definitive operational fit cannot be fully determined.