B.Tech CSE (AI & ML), Class of 2027 · Manav Rachna International Institute of Research & Studies
I build Python applications and explore AI systems that connect models, tools and user interfaces. My interests span software engineering, AI engineering and applied machine learning.
At Eko Bharat Ventures, I worked as an AI Automation Intern on a CSP Account Issue Communication & Tracking Platform: document ingestion, OCR, human-reviewed records, controlled customer outreach and case tracking. The experience taught me to design around privacy, constrained hardware and reliability.
| Project | What to explore | Stack |
|---|---|---|
| Aura Companion | Desktop interface, tool runtime, native workers and MCP integration | React · Electron · Python |
| Aura Autonomous Agent | Observe–reason–act–verify loop, browser tools and service endpoints | Python · FastAPI · Playwright |
| Aura Final-Year Project | Voice interaction, provider adapters and architecture documentation | React · Electron · Python |
| ARIA | Local LLM command planning and a Python desktop interface | Python · Ollama |
| Research Paper | Local/cloud LLM desktop automation manuscripts and sources | LaTeX · Research |
Software: Python, SQL, React/Electron, REST APIs, SQLite, Git, debugging and modular application design.
AI applications: LLM APIs, tool calling, agent workflows, local inference, OCR, voice interfaces and model/provider integration.
Continuing development: machine learning, NLP, evaluation and data-processing fundamentals.
Projects, internships and hackathons have taught me a lot: how to investigate failures, work within hardware limits, validate outputs and explain engineering tradeoffs. I use AI coding tools while reviewing and testing the resulting work.
- Microsoft Certified: Azure Data Scientist Associate — earned May 2026.
- Microsoft Azure AI Fundamentals (AI-900) — score 945/1000.
- Oracle Cloud Infrastructure 2025 Certified AI Foundations Associate.
- Additional learning in NLP, Azure ML, networking and Python.
Participated in the RBIH National Fraud Prevention Challenge. Our team cleared round one; the next stage highlighted the importance of planning data workloads around memory and compute capacity.
Original Aura · Mark LI fork · Mark LI archive · Mark LIV study · MYRAA distribution
Mark repositories preserve FatihMakes upstream authorship and licenses. Archived snapshots and installer-only distributions are identified as such in their READMEs.
Open to internships and entry-level SDE, AI Engineer and AI/ML Engineer opportunities.