AI/ML Engineer with 5+ Years of Experience

Suraj Bilgi

AI/ML Engineer

• **************@*****.*** • +1-551-***-**** • LinkedIn • GitHub • Portfolio SUMMARY

• AI/ML Engineer with 5 years of experience in designing, deploying, and scaling end-to-end machine learning solutions across NLP, Generative AI

(LLMs), Computer Vision, and advanced analytics using deep learning and statistical modeling techniques.

• Expert in building supervised and unsupervised models using Linear/Logistic Regression, Naive Bayes, SVM, KNN, Decision Trees, K-means, Random Forest, Bagging, and Gradient Boosting for predictive analytics and pattern recognition.

• Emerging focus on multimodal AI and scalable AI agents, including Vision-Language Models (CLIP, LLaVA, BLIP), Diffusion Models (Stable Diffusion, ControlNet), Model Context Protocol, vector databases, and guardrails frameworks for safe, production-grade GenAI systems.

• Advanced Python programming and MLOps expertise, end-to-end AI pipelines with feature engineering, data preprocessing, and visualization and deploying production-grade models on AWS using CI/CD, Docker, Kubernetes, and MLflow with robust monitoring, versioning, and reproducibility. SKILLS

Language and Frameworks: Python, MATLAB, Tensorflow, Pytorch, Scikit-learn, Seaborn, Keras, OpenCV, Detectron2, MMDetection, Mediapipe Machine Learning: Linear Regression, Logistic Regression, Decision Trees, Random Forests, SVM, Naive Bayes, Feature Engineering, Model Evaluation. Deep Learning: CNN, RNN, LSTM, NLP, Large Language Models, LangChain, Hugging Face Transformers (BERT, GPT-3), OpenAI APIs Cloud: AWS (EC2, S3, Lambda, API Gateway, CloudWatch, CodeDeploy), GCP (Vertex AI, Google Cloud Storage, AI Platform Pipelines) LLM Techniques: Prompt Engineering, Retrieval-Augmented Generation (RAG), Re-ranking, Knowledge Graph Integration, Fine-tuning (SFT, PEFT, LoRA), Reinforcement Learning with Human Feedback (RLHF) Computer Vision Techniques: Transfer Learning, Object Detection(YOLO, Faster R-CNN,SSD), Semantic Segmentation(U-Net), 3D reconstruction. Database and Tools: SQL Server, MySQL, PostgreSQL, Redis, Neo4j, Apache Airflow, MLflow, Docker, CI/CD(Github Actions), Vector Database Applications: AI Agents, Workflow Automation, Conversational AI, Enterprise Search, Document Summarization, Semantic Search, UGC Moderation, Recommendation Systems, OCR & Document processing, Face recognition, Scene Understanding, Medical Imaging, Video Analytics, Augmented Reality. Certifications: Generative AI Engineering with LLMs Specialization, AWS Certified Solutions Architect – Associate PROFESSIONAL EXPERIENCE

Tracker Groups LLC May 2024 – Sept 2025 Albany, NY Machine Learning Engineer

• Engineered a high-fidelity 3D reconstruction pipeline for indoor house environments using LiDAR-derived point cloud data, integrating Open3D and MeshLab for surface reconstruction and semantic segmentation of architectural elements.

• Developed and containerized backend microservices for preprocessing, mesh generation, and rendering using Python, deployed via Docker and orchestrated on AWS ECS; optimized reconstruction throughput by 40%.

• Designed and deployed a FastAPI-based ML inference service to serve XGBoost models in real-time; integrated JWT authentication for secure access and leveraged AWS SQS to handle asynchronous prediction requests, reducing latency by 30% and enabling scalable model consumption in production.

• Executed development of a model dashboard using Python (Flask), integrated with MLflow and PostgreSQL to track runs, hyperparameters, and model versions. Enabled collaboration between data science and MLOps teams, improving reproducibility and traceability by 40%.

• Constructed and trained deep learning models using CNNs and RNNs to automate document classification and object detection across enterprise workflows, increasing operational throughput by 40% and enabling scalable AI adoption

• Architected and fine-tuned NLP models using BERT and GPT-3 via Hugging Face Transformers for sentiment analysis and intent detection, improving customer feedback classification accuracy by 20% and reducing escalation cycles Ineuron Intelligence Private Limited Jul 2022 – Aug 2023 Bangalore, India AI/ML Developer

• Architected a CNN-based image classification pipeline using TensorFlow and OpenCV to detect diabetic retinopathy in high-resolution retinal scans, achieving 91% AUC and improving early-stage detection in nationwide trials.

• Crafted a scalable NLP workflow using spaCy and regex-based extractors to mine clinical symptoms from unstructured EHR text, leading to a 47% increase in structured patient data and enhancing downstream analytics.

• Formed a Random Forest-based risk prediction model to identify patients likely to develop adverse drug reactions during post-trial follow-ups, reducing clinical error rate by 19% over 6 months.

• Developed a multi-agent Generative AI solution using LangChain, integrating a Database Agent, RAG Agent, and Supervisor Agent to intelligently route user queries, improving banking query resolution efficiency by 40%.

• Built a secure NL-to-SQL Database Agent with read-only validation and PostgreSQL integration, enabling automated credit card eligibility checks while ensuring 100% compliance with financial data governance and reducing manual support effort.

• Implemented a RAG-powered knowledge assistant with FAISS-based semantic retrieval over 100+ pages of banking policies and product brochures, delivering instant, explainable insights on card benefits and eligibility, increasing self-service adoption by 30%. Resolute AI Software Private Limited Jan 2020 – Jul 2022 Bangalore, India Machine Learning Engineer

• Formed real-time AI services using OpenCV and LSTMs to detect abnormal user behaviors in video streams, reducing manual monitoring needs by 55% in fraud detection and compliance scenarios

• Migrated legacy ML models to AWS SageMaker with Docker containers and implemented model endpoints using API Gateway and Lambda, reducing deployment time by 50% and enabling seamless scaling

• Led exploratory data analysis and implemented decision tree models using scikit-learn to identify customer churn indicators, enhancing prediction accuracy by 20% and informing targeted retention strategies across multiple business units.

• Established and deployed predictive models leveraging gradient boosting and ensemble methods, achieving 88% forecasting accuracy for sales trends; this enabled inventory planning and reduced stockouts by 15%, saving significant operational costs.

• Fabricated and automated scalable data pipelines for complex feature engineering, data cleaning, and transformation using Python and SQL, improving data processing speed by 35%, which enhanced model training efficiency across multiple projects.

• Applied advanced NLP techniques, including sentiment analysis and topic modeling, on large volumes of customer feedback data, generating insights that drove a 22% increase in customer satisfaction and informed product roadmap prioritization. EDUCATION

Master of Science in Applied Artificial Intelligence Sept 2023 -May 2025 Hoboken, NJ Stevens Institute of Technology, New Jersey, USA

Bachelor of Technology in Electronics and Communication Aug 2015 -June 2019 Bangalore, KA Acharya Institute of Technology, Karnataka, IND

Contact this candidate