█████╗ ██╗ ██╗███████╗██╗ ██╗ █████╗ ████████╗ ██╗ ██╗███████╗██████╗ ███╗ ███╗ █████╗
██╔══██╗██║ ██╔╝██╔════╝██║ ██║██╔══██╗╚══██╔══╝ ██║ ██║██╔════╝██╔══██╗████╗ ████║██╔══██╗
███████║█████╔╝ ███████╗███████║███████║ ██║ ██║ ██║█████╗ ██████╔╝██╔████╔██║███████║
██╔══██║██╔═██╗ ╚════██║██╔══██║██╔══██║ ██║ ╚██╗ ██╔╝██╔══╝ ██╔══██╗██║╚██╔╝██║██╔══██║
██║ ██║██║ ██╗███████║██║ ██║██║ ██║ ██║ ╚████╔╝ ███████╗██║ ██║██║ ╚═╝ ██║██║ ██║
╚═╝ ╚═╝╚═╝ ╚═╝╚══════╝╚═╝ ╚═╝╚═╝ ╚═╝ ╚═╝ ╚═══╝ ╚══════╝╚═╝ ╚═╝╚═╝ ╚═╝╚═╝ ╚═╝
class AkshatVerma:
def __init__(self):
self.name = "Akshat Verma"
self.role = "Software Engineer @ Scaler"
self.education = "B.Tech, MNNIT Allahabad (2021–2025)"
self.location = "Bengaluru, Karnataka 🇮🇳"
self.ai_stack = ["FastMCP v3", "LangChain", "RAG Pipelines",
"Whisper", "Hugging Face", "Reinforcement Learning",
"Claude Opus 4.6", "Structured Outputs"]
self.backend = ["Spring Boot", "FastAPI", "Django",
"Spring Security 6.5", "JWT Auth", "OAuth2"]
self.infra = ["Docker", "Kubernetes", "Azure", "AWS",
"CI/CD", "PostgreSQL", "Redis", "Kafka"]
self.competitive = {"LeetCode": "Knight 1868 | 850+ problems",
"CodeChef": "3★ | Rating 1688",
"ICPC" : "AIR 707, India Prelims '23"}
def current_focus(self):
return ["AI Agent Orchestration with FastMCP",
"RL workflows for LLM evaluation",
"Scalable Backend Systems",
"Open Source Contributions"]| Category | Technologies |
|---|---|
| Agent Frameworks |   |
| LLMs & Models |    |
| Techniques |    |
| ML Libraries |   |
| Web Crawling |   |
🟢 Scaler — Software Engineer (Dec 2025 – Present, Bengaluru)
- Designed a custom FastMCP v3.0.0 Server to orchestrate tool calls, automate evaluation logic, and manage state-driven task execution
- Implemented async spiders, fetchers & parsers in the Scrapling framework for scalable web crawling
- Resolved Kubernetes & BiomeJS issues; built coding RL workflows validated on Claude Opus 4.6
- Created an Excel-like app using Next.js, FastAPI & SQLite with full persistent state, undo/redo, filters, sorting, and context-aware toolbars
🔵 Intellimark AI — Software Developer (Sep 2025 – Dec 2025, Bengaluru)
- Engineered a Forecasting API with FastAPI, pmdarima, pandas & NumPy for monthly sales prediction
- Built interactive forecasting dashboard with Python 3.12, Django & React.js
- Deployed Dockerized services on Azure with Kubernetes enabling full CI/CD pipeline
🟡 Sunstone — Software Engineer Intern (Jun 2025 – Aug 2025, Gurugram)
- Built backend services with Java 21 & Spring Boot, containerized on AWS with Docker
- Integrated Vimeo API resumable upload via Feign Client with Tus-based video ingestion
- Implemented Google Sign-In (OAuth2), Spring Security 6.5 JWT auth, and CRUD on Google Drive & Calendar APIs
| Project | Stack | Highlights |
|---|---|---|
| ManoHari — E-Commerce | Java 17, Spring Boot, React.js | RESTful APIs, MVC, H2 DB, stock-based cart logic |
| CyberCure — Hackathon Winner | Django, PyTorch, TensorFlow, Streamlit | ResNet50 image forgery, Random Forest IDS, Android malware detection |
| Platform | Achievement |
|---|---|
| 🟡 LeetCode | Knight · Rating 1868 · Global Rank 715 · 850+ problems solved |
| 🟤 CodeChef | 3★ · Rating 1688 · Global Rank 198 |
| 🌍 ACM ICPC | AIR 707 — India Preliminary 2023 |
| 🏅 Hack36 6.0 | 5th Rank — National Hackathon @ MNNIT Allahabad |
| ☁️ Cloud & Open Source | Hacktoberfest · Google Cloud Study Jams |
"Code is like humor. When you have to explain it, it's bad." — Cory House
⭐ Star repos you find useful | 🤝 Open to collaborations on AI & Backend projects