Full-Stack Developer building web apps and AI solutions.
New to Contra
Full-Stack Developer building web apps and AI solutions.
Lira is an AI-powered knowledge assistance platform that helps users find reliable answers from internal documents using Large Language Models (LLMs) and Retrieval-Augmented Generation (RAG).
As a Full-Stack Developer at NLL Academy, I contributed to building and integrating features across the platform, combining modern frontend experiences with backend APIs and AI-powered workflows.
Key contributions:
- Developed responsive user interfaces using React.js, TypeScript, and Tailwind CSS.
- Integrated frontend components with backend APIs built using Python and FastAPI.
- Contributed to document-based question-answering workflows powered by LLMs and RAG.
- Worked on guided learning experiences, learning paths, and conversational interactions.
- Contributed to analytics dashboards and data visualizations for monitoring platform usage.
- Collaborated on debugging, feature integration, and improving overall application usability.
Tech stack: React.js, TypeScript, Tailwind CSS, Python, FastAPI, REST APIs, LLMs, RAG, and SQL.
The project gave me practical experience building AI-enabled software that combines document retrieval, conversational interfaces, and full-stack application development.
0
20
SOP Automation — AI-Assisted SOP Generation
SOP Automation is a document-driven application designed to streamline the creation and management of Standard Operating Procedures (SOPs). It helps structure SOP content into predefined sections, process source documents, and support a more organized drafting and review workflow.
As a Full-Stack Developer at NLL Academy, I contribute to developing the application’s frontend, backend APIs, document-processing workflows, and database integrations.
Key contributions:
-Developed and integrated frontend features using React.js and TypeScript.
-Built and maintained backend APIs using Python and FastAPI.
-Worked on document-processing workflows for extracting and organizing content from Word documents and other supported formats.
-Implemented SOP section mapping and integration with predefined structure definitions.
-Worked on draft state management and saving generated SOP content.
-Contributed to SME guidance features that support the refinement of SOP drafts.
-Troubleshot API, database, and document-processing issues to improve application reliability.
Tech stack: React.js, TypeScript, Python, FastAPI, PostgreSQL, SQLAlchemy, and Docling.
The project has given me practical experience building document-centric business applications, integrating complex backend workflows, and developing tools that help streamline structured business processes.
AI/ML engineer building production RAG & LLM agent systems —
New to Contra
AI/ML engineer building production RAG & LLM agent systems —
Title: NEXUS AI — M&A Intelligence Platform Description: Enterprise platform combining RAG, Neo4j knowledge graphs, and XGBoost ESG forecasting under a LangGraph agent. Built for PE due diligence. Image: same NEXUS AI screenshot from Upwork Link: private demo only
0
44
Title: ASPIRE — Automated ESG Report Generator (GRI/TCFD)
Description: Built a GenAI system that converts raw ESG CSV data into a fully structured GRI/TCFD-aligned sustainability report in ~30 seconds using LLaMA 3.3. The validation layer cross-checks every LLM-generated number against source data before it reaches the report, preventing hallucinated figures in a compliance document. Includes red-flag triggers for fatalities, data breaches, and high turnover that auto-apply GRI materiality language. 57/57 tests passing, deployed on Render, Streamlit Cloud, and AWS EC2.
0
58
Title: LEXINTEL — Autonomous M&A Due Diligence Accelerator
Description:Built a 4-pass autonomous AI pipeline that ingests an entire M&A data room — PDFs, contracts, exhibits — and produces a structured legal report in under 2 minutes. Detects missing exhibits, extracts clauses across 9 legal categories, and catches cross-document contradictions like mismatched equity percentages automatically. Zero data retention, fully ephemeral processing. 0% failure rate in testing. Built with Groq LLaMA 3.3, FastAPI, Streamlit, deployed on AWS EC2.
0
56
Title: NEXUS AI — M&A Intelligence Platform
Description: Enterprise platform combining RAG, Neo4j knowledge graphs, and XGBoost ESG forecasting under a LangGraph agent. Built for PE due diligence.
Image: same NEXUS AI screenshot from Upwork
Link: private demo only
Eeveon Watchtower — Infrastructure monitoring and operational visibility for deployed services. I built the backend monitoring workflow, health checks, alerting logic, and deployment tooling to help teams detect failures early and keep production systems reliable. https://eeveon.com/products
Eeveon Minutes — AI meeting intelligence that turns conversations into structured summaries, decisions, action items, and searchable records. I built the backend and AI workflow for transcription processing, summarization, and reliable product delivery. https://eeveon.com/products/meetings
0
32
Eeveon Deploy — Developer-first CI/CD built with Python and Docker. I designed the deployment workflow, backend automation, packaging, documentation, and PyPI delivery to make releases predictable without vendor lock-in or complex YAML. https://eeveon.com/products/deploy