// devarshi.config.ts
export default {
name: "Devarshi Lalani",
role: ["AI/ML Engineer", "Full-Stack Engineer"],
currently: "Software Development Engineer @ Acqurie.io",
based_in: "Ahmedabad, Gujarat, IN",
timezone: "UTC+05:30",
education: "B.E. Computer Science, L J University (2022–2026)",
} as const;📄 about.md |
🧠 skills.json |
💼 experience.log |
🚀 projects/ |
🎓 education.yml |
|---|
devarshi@portfolio:~$ whoamiAI Engineer with hands-on experience building LLM-powered applications, agentic workflows, and AI orchestration platforms — backed by a full-stack engineering foundation. I work across the stack: LangChain/LangGraph pipelines on one end, React/Node/FastAPI products on the other.
* mar 2026 — present Software Development Engineer @ Acqurie.io
| Built agentic AI workflows & orchestration features (LangChain, LangGraph, RAG)
| for a multi-tenant platform. Designed an LLM usage-metering & credit system
| to track/bill consumption per tenant. Shipped supporting FastAPI + React
| product features, incl. Stripe subscription integration.
|
* jan 2026 — mar 2026 Machine Learning Intern @ fxis.ai
| Applied supervised & unsupervised ML algorithms; hands-on ML/DL
| experimentation workflows.
|
* feb 2025 — jun 2025 Machine Learning Intern @ Feynn Labs
| Designed AI agents automating market segmentation — 40% faster prospect ID.
| Built a regression-based financial health optimizer (95% accuracy) forecasting
| SME cash flow.
projects/
├── AllotIQ/ agentic IPO intelligence platform
│ ├── ai-service/ LangChain + GPT-4o-mini on FastAPI
│ │ → analyst briefs, cited reports, grounded chat
│ ├── scoring-engine/ deterministic apply/wait/avoid signal engine
│ ├── guardrails/ citation validation + hallucination checks
│ └── deployed → allotiq.in (Railway + Vercel + PostgreSQL/Prisma)
│
├── mlcli-toolkit/ open-source ML/DL training CLI (pip package)
│ └── model-registry/ Scikit-learn + TensorFlow (DNN/CNN/RNN/LSTM/GRU)
│
├── AlphaVista/ stock analysis platform
│ ├── prediction/ TensorFlow + Scikit-learn price prediction
│ ├── sentiment/ VADER-based NLP sentiment analysis
│ └── data-layer/ live market APIs + Redis caching
│
└── blockchain-tx-ui/ open-source npm package (TypeScript + React)
└── components/ wallet connection + live tx status tracking
university: L J University, Ahmedabad, India
degree: B.E. Computer Science and Technology
duration: Nov 2022 – Sept 2026
certifications:
- name: MLOps for Generative AI
issuer: Google Cloud
date: April 2025
- name: AWS Cloud Technical Essentials
issuer: Coursera
date: March 2025
- name: Exploratory Data Analysis for Machine Learning
date: January 2025devarshi@portfolio:~$ connect --with devarshi
> process exited with code 0
{ "generative_ai": ["OpenAI API", "Gemini", "HuggingFace", "LangChain", "LangGraph", "RAG", "Prompt Engineering"], "vector_stores": ["FAISS"], "machine_learning": ["Scikit-learn", "Regression", "Classification", "Clustering", "NLP", "Embeddings", "VADER"], "deep_learning": ["TensorFlow", "ANN", "CNN", "RNN/GRU", "LSTM"], "languages": ["Python", "TypeScript", "JavaScript", "Java", "C++"], "frontend": ["React.js", "Tailwind CSS"], "backend": ["Node.js", "FastAPI", "Django", "REST APIs"], "data": ["PostgreSQL", "Prisma ORM", "Redis", "Pandas", "NumPy", "SQL"], "cloud_devops": ["AWS", "Google Cloud Vertex AI", "Vercel", "Railway", "Docker", "Git"] }