Oct 2024 – Mar 2025 · 5 mos · Remote · Remote

• Conducted prompt-based evaluations of large language models (LLMs) across diverse tasks to assess performance, coherence, and failure points.
• Designed challenging, multi-turn, and adversarial prompts to probe model behavior under edge cases and identify vulnerabilities.
• Systematically evaluated and documented model responses for factual accuracy, logical consistency, and alignment with task expectations.
• Collaborated with research teams to iterate on model behaviors by highlighting failure scenarios and providing feedback for fine-tuning.

Nov 2023 – Oct 2024 · 11 mos · Kochi, Kerala, India · On-site

• Mac Team:
• Designed and integrated advanced Mac widget controls for lock screen
• and desktop management features, directly contributing to a 30%
• decrease in unauthorized access incidents, enhancing data security
• across the organization.
• Windows Team:
• Launched a comprehensive enhancement project for firewall features,
• resulting in a 29% increase in system reliability, as measured by uptime
• statistics, while ensuring compliance with industry security standards
• and protocols.
• Developed a Windows policy feature to restrict media usage, controlling media types, user operations, and roles.
• Created a control policy for Windows Store apps, enabling app installation and functionality under predefined rules, either through separate rules or direct app inclusion.
• Developed a real-time Policy Enforcement Engine for UEM using Java, Spring Boot, Kafka, PostgreSQL,
• Docker, and Kubernetes. Ensured instant policy enforcement, secure APIs (JWT, RBAC), 95% test coverage,
• and 40% fewer false alerts with the ELK Stack, deployed for scalability via Kubernetes.
• Built a Remote Device Management & Patch Deployment System using Spring Boot, RabbitMQ,
• TimescaleDB, Prometheus, and Grafana. Enabled secure, asynchronous patch updates, optimized
• deployments 35% faster, and ensured high availability with Kubernetes & Docker while monitoring success rates via Grafana.
• Built a Secure Payment Gateway for UEM using Spring Boot, enabling enterprises to enforce payment
• policies on corporate and BYOD devices. Implemented PCI-DSS tokenization, HMAC authentication, and
• MFA, supporting credit/debit cards, UPI, and corporate wallets for secure in-app transactions.
• Built a real-time policy enforcement engine using Java, Spring Boot, Kafka, and Redis to restrict
• transactions based on device compliance, geolocation, and user hierarchy, ensuring instant enforcement
• and secure, policy-driven payments.

Oct 2022 – Oct 2023 · 1 yr · Bengaluru, Karnataka, India · On-site

• Enhanced reachability bot's functionality using JavaScript​: Added
• MySQL database for queries and integrated Timescale DB for
• improved alerting, employing thorough modifications and
• seamless data integration.
• Created Python-based REST API to enhance Athena service
• efficiency, enabling parallel execution of GET, PATCH, and POST
• requests. Implemented rate limiting and robust handling of failure
• cases and status codes.
• Enhanced shell scripts for parallel testing (alpha, beta) with 67%
• reduced runtime. Ensured compatibility, integrated into Jenkins
• pipelines, and covered edge cases for comprehensive testing.
• Built an Automated Meeting Transcription & Summarization Service using Java (Spring Boot), Python
• (AI/ML), RabbitMQ, AWS S3, and Elasticsearch. The system captures audio, transcribes it with AI models,
• summarizes key points, and indexes transcripts in Elasticsearch, improving efficiency by 40% with RabbitMQ for asynchronous processing.
• I developed an Intelligent Meeting Insights & Analytics Dashboard using Java (Spring Boot), Kafka, Apache Flink, React.js, PostgreSQL, and Grafana. Kafka Streams process real-time meeting data, Apache Flink enables instant analytics, insights are stored in PostgreSQL, and visualized via React.js & Grafana, providing live engagement trends and speaker analytics.

Jul 2022 – Sep 2022 · 2 mos · New Delhi, Delhi, India · Remote

• ➢ Machine Learning: Applied unsupervised machine learning techniques, specifically the
• KMeans clustering algorithm, to determine optimal cluster count for the Iris dataset.
• ➢ Technical Proficiency: Demonstrated expertise in Python programming, utilizing Jupyter
• Notebook for end-to-end project execution. Proficient in data manipulation (Pandas),
• numerical operations (NumPy), and data visualization (Matplotlib).
• ➢ Data Analysis: Leveraged a strong data science foundation to preprocess, analyze,
• TensorFlow, scikit-learn and visually represent data with Power Bi, enhancing insights
• and decision-making.

Apr 2022 – Jul 2022 · 3 mos · Bengaluru, Karnataka, India · Remote

• Analyzed company data with PowerBI, pinpointed sales bottlenecks,
• proficient in PowerBI and SQL for extraction and cleaning. Designed
• intuitive dashboards, resolving 20% sales drop, enhancing data visibility
• and decision-making.