Hi,
I'm Manikanta Sanjay Veera
You’ve got to start with the customer experience and work back toward the technology—not the other way around.
About
Greetings Viewers !!!
I have completed my Masters in Computer Science @
with a passion for leveraging technology to make a positive impact on our world. My journey into the tech world began with a simple 'Hello World' program and has evolved into a fulfilling career where I’ve built and deployed machine learning models across various industries.
My academic contributions include authoring technical papers presented at IEEE and Springer conferences, covering topics from healthcare to time series forecasting.
I thrive in dynamic environments where quick thinking and adaptability are key. I’m ready to take on roles such as Data Scientist, Computer Vision Engineer, Machine Learning Engineer, Data Engineer, and Data Analyst. My strong foundation in Reinforcement Learning, Natural Language Processing, Data Science, Artificial Intelligence, and Machine Learning equips me well for these challenges.
Outside of work, you’ll find me hiking scenic trails or on the cricket and soccer fields. I’m always on the lookout for new opportunities to apply my skills and passion for technology in impactful ways.
Let’s connect and explore how we can work together to push the boundaries of technology!
Education
Experience
San Jose State University
San Jose, California- Container Caching Optimization for Serverless Edge Computing - Devised custom DQN architecture in C++ for learning resource optimization strategies, minimizing cold startup delays; led to cache hit ratios of 0.83.
- Introduced heuristic algorithm-based action masking to tackle scalability issues across large action spaces (560K).
- Performed quantization (float32 to float16) on TensorRT-based models, increasing CPU/GPU utilization (150%/90%).
- Providing constructive feedback on assignments and full-stack projects for 25+ students, utilizing Agile methodologies, Git, Java, Jenkins, Selenium, Flask; resolve 10+ weekly student inquiries.
- Supporting curriculum for 35+ students, emphasizing CPU architecture, memory management, software-hardware integration through hands-on C++ and assembly language exercises.
- Mentoring 30+ students in blockchain and cryptocurrencies, focusing on the development of smart contracts and decentralized applications (DApps) using Solidity and Rust.
- Graded numerous assignments across technologies, including Solidity, Rust, and JavaScript, evaluating code quality, problem-solving techniques, and the practical application of blockchain concepts.
Arkoz
San Ramon, CA AI Engineer Intern 08/2023 - 12/2023- Initiated transformation of GenAI SEO Copilot app from concept (0) to fully functional platform (1) by integrating RAG models providing recommendations for website optimization; led to 30% enhancement in client retention rates.
- Analyzed Google Search Console data via GPT-4 API, delivering insights on session duration, hits, and bounce rate.
- Executed prompt engineering with GPT-4 to refine user queries, resulting in tailored content optimization suggestions.
- Deployed app on AWS EC2 P3 with Elastic Load Balancing; guaranteed stable performance during peak traffic.
Cognizant
Bengaluru, India- Proposed multivariate time series forecasting model using SoTA hybrid VAEs to predict energy consumption for 130,000+ households, yielding <2.5% MAPE, outperforming Temporal Fusion Transformers by 3x.
- Applied MatrixProfiler(R package) for anomaly detection, identifying 10+ motifs and discords, signaling key trends.
- Executed distributed ML training using JAX, CuML, Dask for 60x accelerated Jupyter workflows on HPC clusters.
- Employed MLflow for A/B testing and experiment monitoring, utilized Optuna for hyperparameter tuning of PyTorch models, enabling seamless performance tracking across 15+ versions.
- Enhanced ml infrastructure with statistical modeling techniques and hypothesis testing in SAS; identified cost-saving opportunities, resulting in 12% reduction in operational expenses, equivalent to $2.5M annually.
- Utilized Bayesian probabilistic modeling to develop predictive churn models, accurately identifying at-risk customers; resulted in 25% reduction in attrition and revenue increase of $500,000.
- Streamlined ingestion of data from 10+ disparate data sources (in CSV, XML, JSON formats) into GCP by transforming data using SQL and R and automated BigQuery schema provisioning with Terraform.
Agent Technologies Inc.
Bengaluru, India Data Scientist 02/2020 - 01/2021- Engineered recommender systems using TensorFlow on AWS SageMaker, designed to enhance user interaction with home automation systems, delivering personalized automation settings with 40% increase in CTR.
- Enhanced pricing strategy by leveraging Markov Chain Monte Carlo methods to conduct market analysis; recommended optimal pricing structure resulting in a 15% increase in revenue and 20% growth in market share.
- Conceptualized and implemented custom ETL process utilizing PL/SQL in Databricks that processed 100M transaction records per day, resulting in storage of 2+ petabytes of data.
- Leveraged Docker to containerize ML deployment, cutting environment setup time from 3 hours to 30 minutes.
Raptor Analytics
Bengaluru, India Machine Learning Intern 09/2019 - 01/2020- Performed semantic segmentation of Left Ventricle in cardiac ultrasound images with floU of 95%.
- Created Diction Fraction prediction model with 5.91 test MAD using R2+1D PyTorch architecture.
- Worked on using Swin Transformers, a Vision Transformer architecture, to denoise cardiac ultrasound images.
Skills
Programming Languages
Databases
Libraries
Platforms
Generative AI
Deployment Tools
Projects
Time Series Analysis on Stocks
Performed EDA on tech giants' stock data, focusing on closing prices and correlations. Created interactive dashboards with real-time data from Yahoo Finance, stored in MongoDB. Implemented financial indicators, MACD, MFI, and RSI, for trading insights.
Number Plate Recognition
A collaborative project associated with Python Application Programming 17CS664 at JSSATE Bengaluru that utilizes image processing and deep learning to detect and recognize vehicle license plates from images. The system offers a interface to allow users to upload vehicle images and view real-time results.
WeHeal
A project built during Stanford Treehacks 2024. RAG-based chatbot powered by Langchain to assist mental health therapists by summarizing patient case files and offer personalized suggestions based on past and ongoing interactions.
Language Proficiency Checker
A project associated with CS286 - Advanced Topics in Natural Language Proficiency Checker at SJSU for grading essays using regression analysis on user-input text. It utilizes fine-tuned LLMs like ROBERTa and DEBERTa v3-small, along with LSTMs, assessing language parameters like cohesion and grammar. Results are displayed via a React-based UI.
Concept Drift Detection
A project associated with CS 271 - Topics In Machine Learning at SJSU focused on evaluating concept drift in malware classifiers using Android datasets, DREBIN and AndroZoo. Employed Adaptive Random Forest (ARF) and Stochastic Gradient Descent (SGD) classifiers for static malware detection. Compared feature extraction performance on textual attributes.
Medicine Supply Chain NFTs
A collaborative project associated with CS 168 - Blockchain and Cryptocurrencies at SJSU on Medicine Supply Chain. Deployed ERC-721 smart contracts on Ethereum and created NFTs for medicines, with metadata hosted on IPFS. Developed a pipeline feature to monitor the flow of products across various stakeholders.