Sadu Shashank Reddy profile photo 1 on Tal

Sadu Shashank Reddy

SDE-2 @ Verizon ( Building High Performance and Scalable Frontend Application React js , typescript, JavaScript(ES6),Redux)

Hyderabad

3+ years of work experience

Verizon
Dec 20232 yrs 5 mosApr 2026

• Architected a robust enterprise AI Assistant utilizing a LangGraph state machine to orchestrate complex Agent-Tool-Agent execution flows, dynamically generating and rendering contextual HTML responses for end-users. • Optimized LLM token consumption and response latency by implementing a sliding-window conversational memory (retaining the last 5 user turns) via Redis, seamlessly managing context window limits during extended interactions. • Developed a Retrieval-Augmented Generation (RAG) pipeline integrated with a vector database for precise technical glossary lookups, maximizing retrieval accuracy and relevance through cosine similarity scoring. • Engineered a highly resilient unified agent service featuring custom tool wrappers with automated retry mechanisms, deterministic fallback protocols, and active rate-limit logging to gracefully handle API timeouts and mitigate LLM hallucinations. • Enforced strict data privacy and security standards by implementing pre-LLM PII redaction pipelines, utilizing the nh3 library for robust HTML sanitization, and configuring secure 5-minute session timeouts. • Decoupled LLM system prompts from the core codebase into a dedicated, version-controlled configuration service, enabling zero-downtime prompt updates and rapid iteration of AI personas. • Integrated a multi-layered logging and observability suite with Prometheus and Grafana, enabling real-time monitoring of LLM workflows, detailed tracing of user-agent-tool interactions, and rapid identification of production anomalies.
Verizon
Jun 20221 yr 1 moJun 2023

• Architected a customizable drag-and-drop dashboard for technicians and supervisors using React-Grid-Layout, supporting dynamic layout configurations across 5 responsive breakpoints. • Engineered high-performance rendering by offloading heavy layout space optimization algorithms to Web Workers, preventing main-thread blocking during complex drag-and-drop operations. • Optimized React component lifecycle utilizing useMemo, useCallback, and useRef to memoize expensive calculations within data-heavy widgets (Job Status, Schedule Management, Interaction Matrix). • Enforced strict engineering standards across the project utilizing TypeScript for type safety, Jest for unit testing, and a custom Webpack/Babel build pipeline. • Integrated robust API communication between front-end dashboards and backend microservices using Axios with advanced error-handling strategies and request debouncing, enhancing data reliability and application responsiveness. • Facilitated troubleshooting processes by leveraging Redux DevTools and Chrome Developer Tools to diagnose, isolate, and resolve 30+ recurring UI and API integration issues monthly, reducing bug resolution turnaround by 45% over 12 months.
Verizon
Feb 20225 mosJun 2022

Hyderabad, Telangana, India

• Architected a live dispatch mapping interface using React-Leaflet to simultaneously visualize multi-layered geospatial data, including dynamically fetched facility hubs, active job sites, and real-time technician coordinates. • Integrated asynchronous REST APIs to plot job and facility locations on an interactive Leaflet canvas, utilizing React state management to efficiently map payload coordinates to interactive UI markers. • Implemented active geolocation state tracking to capture and continuously render the technician’s physical location in real-time in relation to assigned jobs and facilities. • Designed a contextual map UI featuring customized visual distinction (via custom SVG icons and dynamic popups) between jobs, facilities, and personnel, significantly reducing visual clutter for the end-user. • Automated geospatial data updates by scripting backend processes that synchronize fresh job, facility, and technician information to the front end, streamlining real-time mapping workflows and increasing data accuracy for dispatch operations.

Education

B. tech, Computer Science

Sathyabama University

20184 yrs2022

SPR Residential School Yellapur

20133 yrs2016

Join Sadu and 10k+ other professionals on

Tal Boss

Tal is the app where bosses meet engineers & builders directly, without middlemen

Download Tal Boss

Sadu Shashank Reddy on Tal

Headline

SDE-2 @ Verizon ( Building High Performance and Scalable Frontend Application React js , typescript, JavaScript(ES6),Redux)

Location

Hyderabad

Work experience

Software development engineer 2 at Verizon, Dec 2023 - Apr 2026 · 2 yrs 5 mos, • Architected a robust enterprise AI Assistant utilizing a LangGraph state machine to orchestrate complex Agent-Tool-Agent execution flows, dynamically generating and rendering contextual HTML responses for end-users. • Optimized LLM token consumption and response latency by implementing a sliding-window conversational memory (retaining the last 5 user turns) via Redis, seamlessly managing context window limits during extended interactions. • Developed a Retrieval-Augmented Generation (RAG) pipeline integrated with a vector database for precise technical glossary lookups, maximizing retrieval accuracy and relevance through cosine similarity scoring. • Engineered a highly resilient unified agent service featuring custom tool wrappers with automated retry mechanisms, deterministic fallback protocols, and active rate-limit logging to gracefully handle API timeouts and mitigate LLM hallucinations. • Enforced strict data privacy and security standards by implementing pre-LLM PII redaction pipelines, utilizing the nh3 library for robust HTML sanitization, and configuring secure 5-minute session timeouts. • Decoupled LLM system prompts from the core codebase into a dedicated, version-controlled configuration service, enabling zero-downtime prompt updates and rapid iteration of AI personas. • Integrated a multi-layered logging and observability suite with Prometheus and Grafana, enabling real-time monitoring of LLM workflows, detailed tracing of user-agent-tool interactions, and rapid identification of production anomalies.

Software Engineer at Verizon, Jun 2022 - Jun 2023 · 1 yr 1 mo, • Architected a customizable drag-and-drop dashboard for technicians and supervisors using React-Grid-Layout, supporting dynamic layout configurations across 5 responsive breakpoints. • Engineered high-performance rendering by offloading heavy layout space optimization algorithms to Web Workers, preventing main-thread blocking during complex drag-and-drop operations. • Optimized React component lifecycle utilizing useMemo, useCallback, and useRef to memoize expensive calculations within data-heavy widgets (Job Status, Schedule Management, Interaction Matrix). • Enforced strict engineering standards across the project utilizing TypeScript for type safety, Jest for unit testing, and a custom Webpack/Babel build pipeline. • Integrated robust API communication between front-end dashboards and backend microservices using Axios with advanced error-handling strategies and request debouncing, enhancing data reliability and application responsiveness. • Facilitated troubleshooting processes by leveraging Redux DevTools and Chrome Developer Tools to diagnose, isolate, and resolve 30+ recurring UI and API integration issues monthly, reducing bug resolution turnaround by 45% over 12 months.

Intern at Verizon, Feb 2022 - Jun 2022 · 5 mos, Hyderabad, Telangana, India, • Architected a live dispatch mapping interface using React-Leaflet to simultaneously visualize multi-layered geospatial data, including dynamically fetched facility hubs, active job sites, and real-time technician coordinates. • Integrated asynchronous REST APIs to plot job and facility locations on an interactive Leaflet canvas, utilizing React state management to efficiently map payload coordinates to interactive UI markers. • Implemented active geolocation state tracking to capture and continuously render the technician’s physical location in real-time in relation to assigned jobs and facilities. • Designed a contextual map UI featuring customized visual distinction (via custom SVG icons and dynamic popups) between jobs, facilities, and personnel, significantly reducing visual clutter for the end-user. • Automated geospatial data updates by scripting backend processes that synchronize fresh job, facility, and technician information to the front end, streamlining real-time mapping workflows and increasing data accuracy for dispatch operations.

Education

B. tech, Computer Science, Sathyabama University

SPR Residential School Yellapur