Abhishek Goyat

Generative AI Engineer with 3.7 years of specialized experience in Python-driven Deep Learning and Generative AI. Expert in architecting complex, stateful multi-agent workflows using LangGraph and LangChain, delivering a 40% increase in task automation accuracy over linear RAG pipelines. Proficient in the full LLM lifecycle — from PyTorch-based model development to advanced fine-tuning using LoRA/QLoRA and 4-bit quantization (BitsAndBytes).

Proven track record of reducing model memory footprints by 60% while maintaining 95% performance benchmarks for production-scale deployment.

Generative AI / ML Engineer at Azcom Technology (2022-07 – 2026-02)

AI-Driven 5G Network Validation & Intelligent Testing Framework

• Built a modular, AI-augmented validation framework for 3GPP-compliant 5G workflows, enabling seamless interoperability across UE, gNB, 5G Core, and application layers — improving system observability and debugging efficiency.
• Developed AI-assisted test orchestration using pattern recognition and historical test data to prioritize high-risk scenarios, reducing redundant executions and improving fault detection speed.
• Applied anomaly detection on network logs and E2E traces to proactively identify failures and edge-case behaviors, minimizing late-stage defects and improving system stability.
• Achieved ~80% automation coverage of critical workflows, cutting manual validation effort by more than half and enabling faster iteration cycles aligned with CI/CD pipelines.
• Reduced repetitive lab effort by ~35%, enabling teams to focus on root cause analysis, model-driven insights, and performance optimization.

Bachelor of Technology in Information Technology – JC Bose University of Science and Technology, YMCA