Developed and integrated EAGLE-based Bigram Token Processing within the ai-dynamo/dynamo repository, enhancing the KV event reporting system to support overlapping token windows for improved token sequence handling and accuracy. The work involved aligning new processing logic with existing KV event reporting components through targeted compatibility adjustments, reducing integration risk and ensuring system stability. Addressed a bug in Eagle bigram token handling, further stabilizing the tokenization pipeline. Demonstrated proficiency in Python, Rust, and asynchronous programming while collaborating with cross-team contributors. The resulting improvements increased data fidelity for analytics and downstream decision-making, and strengthened the reliability of backend event reporting workflows.