Machine Learning Engineer

LIKITHA TADAKALA

+1-317-***-**** ***************@*****.*** LinkedIn GitHub

PROFESSIONAL SUMMARY

Engineer with 2+ years of production experience across Machine Learning, Data Science, and Software Engineering; shipped YOLO and VLM pipelines at 90 to 95 percent mAP and a data cleaner 63x faster than T5. AWS Certified. Strong in Python, AWS, SQL, Power BI.

TECHNICAL SKILLS

Languages and Tools: Python, SQL, R, JavaScript, Git, GitHub, Docker, Linux, Bash, REST APIs ML and Deep Learning: PyTorch, TensorFlow, scikit-learn, Hugging Face Transformers, Transfer Learning, Fine Tuning, Model Evaluation, Feature Engineering, Explainable AI LLMs and Generative AI: RAG, Vision Language Models, LangChain, Ollama, Mistral, LLaMA2, Google Gemini, FAISS, Prompt Engineering, Offline Inference

Cloud and MLOps: AWS (SageMaker, Lambda, S3, EC2, IAM, Rekognition, Textract), FastAPI, Flask, Model Deployment, Batch Inference, CI/CD, Model Monitoring Data and Analytics: Pandas, NumPy, EDA, Statistical Modeling, Predictive Modeling, Power BI (DAX), Tableau, Matplotlib, Seaborn

Computer Vision and Certs: YOLO, DETR, OpenCV, Object Detection, Image Segmentation, OCR PROFESSIONAL EXPERIENCE

Actual Reality Technologies Data Scientist Perrysburg, OH Sep 2025 to Present

• Built and shipped an end to end ML inference pipeline integrating YOLOv11-X, DETR, Google Gemini 2.5 Flash VLM, and AWS Textract OCR on 700+ engineering drawings at 2560px, hitting 90 to 95 percent mAP@0.5:0.95, eliminating 30 to 40 percent rebar weight estimation errors, and cutting manual processing time 70 percent.

• Orchestrated AWS ML infrastructure for scalable batch inference; resolved 37 to 1.7 percent class imbalance via synthetic data generation; built Power BI dashboards for KPI monitoring; co-authored arXiv paper. CodesOnBytes Jr Data Scientist Remote Dec 2023 to Jan 2024

• Built a Python (Pandas) data ingestion pipeline to harvest and clean raw data from public REST APIs; developed a linear regression forecasting model in scikit-learn and conducted EDA with Matplotlib and Seaborn to identify outliers and feature correlations before model training. Ozibook Data Analyst Remote Mar 2023 to Dec 2023

• Designed and deployed interactive Power BI dashboards visualizing key business metrics, reducing weekly reporting cycle time and giving leadership real time KPI visibility; partnered with stakeholders to define metrics and validate data quality.

Intel Corporation Industrial Trainee Remote Apr 2023 to Jul 2023

• Fine tuned YOLOv5 in PyTorch on a Roboflow automotive dataset, achieving 90 percent mAP@0.5 across 3 classes; applied focal loss and augmentation to resolve class imbalance and calibrated the accuracy and latency trade off for real time inference.

PROJECTS

SmartClean AI Hybrid ML Framework for Automated Data Cleaning

• 63x faster than Hugging Face T5 (33ms vs 2s per 1,000 records), 100 percent format consistency vs T5's 90.3 percent, and 98 percent quality score; deployed as a Flask REST API with zero API costs. YouTube Learning Chatbot Privacy Preserving RAG Pipeline

• Built an offline RAG system using FastAPI, Gradio, FAISS, and locally hosted LLMs (Mistral, LLaMA2 via Ollama) with transcript chunking for unstructured video content and zero external data transmission. Vision For The Blind Edge AI Assistive System on Raspberry Pi 4

• Deployed Inception V3 (97.3 percent accuracy, 124 classes via transfer learning) and Tesseract OCR on Raspberry Pi 4 with a frame deletion memory pipeline for sub second audio feedback on edge hardware. PUBLICATIONS AND CERTIFICATIONS

• Beyond Visualization: Building Decision Intelligence Through Iterative Dashboard Refinement, arXiv, 2025.

• VISION: Visual Insight Solution Interface Outreach Navigator, IEEE Xplore.

• AWS Certified AI Practitioner, AWS Certified Cloud Practitioner EDUCATION

University of Toledo MS Computer Science (AI) Aug 2024 to May 2026 GPA: 4.0 / 4.0 Malla Reddy College of Engineering B.Tech CSE (Data Science) Aug 2020 to May 2024 GPA: 3.6 / 4.0

Contact this candidate