Experience

• 

  Bosch global software technologies

  Jun 2022 - Jul 2022

  Machine learning engineer

  • Developed a proprietary dataset tailored for analyzing Random Telegraph Noise (RTN) in automotive sensor data, enabling more precise modeling of transient signal anomalies.• Applied advanced signal and image processing techniques in combination with DBSCAN clustering to detect RTN-induced outliers, achieving over 90% detection accuracy in benchmark test scenarios.• Engineered a modular RTN detection algorithm, encapsulated in a version-controlled YAML configuration for deployment flexibility and reproducibility across systems.• Conducted exploratory data analysis to differentiate between benign signal fluctuations and RTN events, using both statistical and visual analysis techniques.• Collaborated with system engineers and domain experts to validate findings and refine detection thresholds based on sensor-specific behavior.• Documented the entire pipeline, from dataset construction to RTN classification, ensuring future extensibility and team onboarding. Show less

• 

  Bosch global software technologies

  Feb 2023 - May 2023

  Machine learning engineer

  • Translated MATLAB-based clustering algorithms into scalable PySpark pipelines to classify and detect defective automotive sensors. • Leveraged distributed computing to optimize anomaly detection performance across large-scale sensor datasets.• Automated defect classification workflows, reducing manual inspection time by 60% and increasing operational efficiency.• Tuned machine learning models using hyperparameter optimization, improving model accuracy and reducing training time by 35%.• Visualized sensor behavior and clustering results using Seaborn and Matplotlib, enabling deeper insight into failure patterns.• Collaborated cross-functionally with engineers and data scientists to align model outputs with diagnostic use cases.• Shared ML best practices and reusable code templates, promoting team-wide adoption of scalable analytics techniques. Show less

• 

  Chain-sys corporation

  Jul 2024 - now

  Data analyst

  Responsibilities:• Designed and deployed a cutting-edge NLP-to-SQL LLM, enabling users to convert natural language into accurate SQL queries, reducing manual query writing by 80%.• Fine-tuned large language models using domain-specific data; trained on over 1,000+ annotated SQL examples for increased accuracy in intent recognition.• Built internal tools to benchmark model performance using metrics like BLEU score, execution accuracy, and query latency.• Created a feedback loop for continuous model improvement by integrating user interactions and query outcomes into retraining datasets.• Engineered REST APIs to expose LLM functionalities for integration with internal dashboards and tools.• Implemented version control for training datasets and model checkpoints to ensure traceability and reproducibility.• Collaborated cross-functionally with database admins and data analysts to ensure alignment between LLM outputs and business-critical data queries.• Conducted error analysis and model debugging to resolve edge-case failures and improve overall system reliability.• Explored prompt engineering techniques and model configuration tuning to optimize LLM performance for multilingual inputs and ambiguous queries.• Developed visual dashboards in SAC (SAP Analytics Cloud) to track LLM usage metrics and system performance in real time.• Participated in weekly sprint planning and Agile ceremonies, contributing to roadmap planning and feature prioritization. Show less