001/Portfolio · 2026
Shaik AbdulHafeez.
I build the thing I wished existed when I hit the problem.
Almost everything here started as a problem I ran into myself. Something annoys me, I decide it ought to exist, and I build my way out of it. I learn far more from that than from any tutorial, and I lean on open source the whole way, so I try to put work back into it where I can. Strophic isn't a company and I'm not its founder. It is the name on my workshop, where the experiments live.
Learn things properly by building them, and keep the ones that survive contact with reality.
What I keep building
AI Integration
Putting modern models inside tools that already exist, to find out where they genuinely help.
Workflow & Process Automation
Cutting out the repetitive, error-prone steps that sit between one system and the next.
Custom Software Development
Building the whole thing end to end, because that is how I learn what actually breaks.
AI Agents & Chatbots
Assistants that do real work: answering, triaging, and acting across systems.
Cloud & API Integrations
Wiring payments, data and infrastructure together, and keeping the joins reliable.
Micro-SaaS Development
Taking an idea from a weekend prototype to something people can actually sign up for.
Rules I keep
Build what I needed
Most of it starts as a problem I hit myself.
Ship the truth
Working demos over polished promises.
Open by default
I lean on open source constantly, and put work back where I can.
Stay curious
Following the new thing far enough to know whether it matters.
Four years of shipping into production.
From responsive marketing sites to enterprise OCR pipelines processing twenty thousand invoices a day. Expand any entry for the full record.
RealPage, Inc.
AI Engineer
Production AI-powered OCR platform for enterprise Spend Management
NestJSReact.jsOCRAI Pipelines- Architected and developed a production-grade AI-powered OCR platform for enterprise Spend Management, owning the complete solution from architecture and development to production deployment.
- Built the complete backend using NestJS and frontend using React.js, delivering a scalable end-to-end enterprise application.
- Designed high-throughput document processing pipelines handling 10,000–20,000 invoices per day at ~98% OCR extraction accuracy.
- Developed intelligent invoice extraction pipelines for vendor details, metadata, line items, taxes, totals, purchase-order and payment information across diverse invoice formats.
- Implemented OCR confidence scoring and validation with automated quality assessment plus manual review workflows for low-confidence fields.
- Designed automated invoice verification and financial reconciliation workflows (subtotal, tax, discount, grand-total validation) to guarantee data consistency.
- Built AI-powered product categorization and line-item classification pipelines mapping products into standardized spend categories for enterprise reporting.
- Optimized throughput, latency, and reliability with asynchronous processing and fault-tolerant workflows.
ZerocodeHR
Senior Software Engineer, AI / Agentic Systems
Self-hosted AI infrastructure, agents, and RAG pipelines with the CTO
LangChainvLLMQdrantFastAPIMCPDocker- Collaborated directly with the CTO to architect self-hosted AI infrastructure: local model inference with vLLM, vector databases for RAG, and robust data pipelines for enterprise AI applications.
- Designed a multifunctional company chatbot for clients, developers, and professional services, integrated with Gmail and Google Chat, and extended Jira via MCP to automate ticket creation, updates, and daily reporting.
- Enhanced the chatbot with intelligent agent features (page-context explanations, guided navigation, in-page data analysis, and chart generation from grid data) with guardrails against toxicity and prompt injection plus rate limiting.
- Built high-accuracy RAG pipelines with LangChain, LlamaIndex, QdrantDB, and GPT-oss, designing custom ingestion, chunking, and retrieval strategies for large-scale document processing.
- Designed advanced AI agents using LangChain and OpenAI Agent SDK for site building, report generation, data modeling, and site automation.
- Built a secure, Docker-isolated Data Analysis Agent processing user-uploaded CSV/Excel with Pandas and Plotly, generating professional LaTeX PDF reports.
- Fine-tuned open-source LLMs (DeepSeek, Qwen) for domain-specific contextual understanding and accuracy.
- Architected high-performance FastAPI backends with Prometheus, Grafana Loki, and MLflow for metrics logging and model tracing.
- Deployed scalable ML model APIs with AWS Lambda, API Gateway, and IAM.
- QA Automation: built frameworks with Java, Selenium, TestNG, and Spring Boot, cut suite execution time by 50% (2h to 1h) and fixed 200+ test cases.
Xyenta Solutions
Full Stack Web Developer, Intern
Fintech prototype with Spring Boot APIs and a Flutter frontend
Spring BootJavaFlutterDart- Developed Spring Boot APIs in Java with DTOs, repositories, and configurations.
- Built Flutter pages in Dart for a fintech prototype; optimized logic for iOS/Android responsiveness.
- Ensured seamless backend-to-frontend integration and performance; collaborated via Git/GitHub.
Codecrux Web Technologies
Web Developer, Intern
Responsive websites and CI/CD fundamentals
HTMLCSSJavaScriptWebflow- Developed responsive websites with HTML, CSS, JavaScript; used Webflow, WordPress, Bootstrap, Tailwind.
- Learned CI/CD workflows to improve deployment efficiency and reliability.
Things I built to find out if they would work.
Every build in one place: the Strophic products, the agents and pipelines, the one-offs. Some became businesses, some became lessons; all of them shipped.
Things other people asked me to build.
Work that came in through Strophic. The same workshop, pointed at somebody else's problem.
Selected engagements
The tools I reach for, and the ones I keep reaching for.
57 technologies across models, pipelines, backends and interfaces, grouped by how often they end up in the repo.
AI, ML & GenAI
- Agentic AI Systems
- RAG Pipelines
- LLM Fine-tuning
- MCP
- Embeddings & Vector Search
- OCR
- Prompt Engineering & Guardrails
- Full-Stack Web Development
- Machine Learning
- NLP
- Testing & Automation
Frameworks & Libraries
- LangChain
- LlamaIndex
- OpenAI Agent SDK
- CrewAI
- Phidata
- vLLM
- Ollama
- React.js
- Next.js
- Node.js
- NestJS
- FastAPI
- Flask
- Spring Boot
- Flutter
- Pandas
- Plotly
Languages
- Python
- Java
- JavaScript
- TypeScript
- Dart
- HTML
- CSS
- SQL
Cloud, Data & Tools
- Qdrant
- ChromaDB
- PostgreSQL
- MySQL
- MongoDB
- Redis
- Firebase
- AWS
- Docker
- Prometheus
- Grafana
- MLflow
- Git
- GitHub
- Figma
- Jira
Design & Craft
- UI/UX Design
- Photoshop
- Video Editing
- Adobe XD
- Canva
Where it was learned, and what got written down.
Education
The ICFAI University For Higher Education
B.Tech in Computer Science Engineering
9.55
Sri Chaitanya Junior College
Math, Physics and Chemistry
9.88
Bhashyam Public School
Secondary Schooling
10
Let’s build something
that ships.
An AI integration, an agentic system, or a product idea that has been sitting in a notes app for six months. I’m happy to talk through any of it. Reach me directly, or take a look at what I’ve been building.