CV
Education, experience, projects, and technical skills.
Contact Information
| Name | Akhil Shekkari |
| Professional Title | AI Engineer |
| shekkari.akhil@gmail.com |
Professional Summary
AI engineer interested in LLM inference, serving, and agent systems.
Experience
-
2025 - 2025 Silver Spring, MD
AI Engineer
Atrium — Pfizer R&D project
- Worked on AI-assisted authoring for clinical-trial statistical analysis plans in a regulated pharmaceutical environment.
- Co-authored a peer-reviewed publication in Clinical Trials.
-
2022 - 2024 India
Machine Learning Engineer
Tezo
- Worked across machine-learning and data-engineering projects.
-
2021 - 2022 India
Junior Machine Learning Engineer
Tezo
- Worked on Snowflake data migration, SQL analytics, and machine-learning workflows.
Education
-
2024 - 2026 College Park, MD
Master of Science
University of Maryland, College Park
Applied Machine Learning
- Advanced Machine Learning, Deep Learning, Natural Language Processing, Probability and Statistics, Optimization
Projects
-
mini-vllm
A small single-GPU inference engine implementing paged KV cache, block allocation, and continuous batching.
- Paged KV cache and a shared fixed-size GPU block pool.
- FCFS continuous batching with separate prefill and decode phases.
-
coding-agent
A plan-then-execute CLI agent for making changes in real Git repositories.
- Repo-scoped read, search, write, and shell tools.
- Self-hosted inference through vLLM with optional LangSmith tracing.
-
AI Agent Framework
A multi-step tool-use agent built from scratch with MCP integrations and a FastAPI backend.
Skills
Languages and libraries: Python, PyTorch, SQL, NumPy, Scikit-learn
LLM systems: KV cache, paged attention, vLLM, SFT, LoRA, QLoRA, Hugging Face Transformers
Agents and APIs: Function calling, MCP, LangGraph, FastAPI
Infrastructure and data: Docker, AWS, Azure ML, ClearML, Snowflake, vector databases