CUDA-Accelerated AI/ML Engineer

Venkata Krishna Vamsi Vipparthi

+91-630******* # ****************@*****.*** ï linkedin § GitHub Portfolio Experience

Unistring Tech Solutions Hyderabad, India

AI/ML Engineer Jan 2024 – Present

– Spearheaded the end-to-end architecture of a real-time signal classification system as the sole technical lead, managing the full ML lifecycle from high-velocity data collection to containerized deployment.

– Architected an asynchronous, multi-threaded orchestration layer routing real-time data streams across 4 dynamic carrier frequency bands based on zero-shot carrier classification.

– Developed production-grade Python services fused with bare-metal CUDA C++ routines, optimizing memory allocation and thread hierarchy to drop execution latencies from > 10s to < 100ms.

– Designed a custom auto-labeling state machine and on-device Continual Learning loop with human-in-the-loop (HITL) verification screens, updating production weights for 100+ target classes with zero downtime. Projects

Omni-Route: CX Multi-Agent Orchestrator June 2026

Time-boxed 16-Hour Engineering Spike for Hybrid LLM Orchestration GitHub LangSmith Trace

– Architected a hybrid multi-agent orchestration layer coordinating state transitions between an internal FAQ retrieval agent and simulated 3rd-party vendor support APIs.

– Implemented zero-shot intent routing via LangGraph, utilizing interrupt before primitives to freeze execution state into SQLite for secure Human-in-the-Loop (HITL) support escalations.

– Integrated end-to-end LangSmith tracing telemetry to observe token latency, prompt execution paths, and intent routing accuracy.

GiftAI - AI-Powered Recommendations May 2025 - June 2025 Full-Stack Application built during the month-long Bolt.new Hackathon GitHub Devpost

– Leveraged Generative AI as a force multiplier to rapidly learn and implement an unfamiliar tech stack (Next.js, TypeScript, Tailwind, Supabase) entirely from scratch within a month.

– Engineered an interactive recommendation platform integrating Google Gemini AI to provide context-aware, multi-currency gift suggestions based on natural language inputs.

– Designed and deployed the complete database schema and authentication flows utilizing PostgreSQL via Supabase. Technical Skills

Languages: Python, C++, CUDA C, SQL

AI & LLM Stack: LangGraph, LangSmith, PyTorch, Scikit-Learn, PyCUDA, NumPy, Pandas Systems & Infrastructure: Linux OS, SQLite, Docker, Git, Multi-Threading, REST APIs Web/Full-Stack: Next.js, React, Tailwind CSS, PostgreSQL (Supabase) Core Competencies: LLM Orchestration, State Machine Design, Production ML Services, Low-Latency Optimization Education

National Institute of Technology, Warangal CGPA: 8.13/10 B.Tech in Biotechnology 2020 – 2024

FIITJEE Junior College Percentage: 94.1%

Board of Intermediate Education, Telangana 2018 – 2020 Bhashyam High School CGPA: 10/10

Board of Secondary Education, Telangana 2017 – 2018 Certifications & Achievements

Applied Machine Learning in Python,Coursera Jun 2023 Microsoft Technology Associate (MTA): Python Programming,Microsoft Second Place,Kaggle Competition on Wild Forest Fire Detection using Deep Learning Mar 2023

Contact this candidate