Developer building practical systems across AI workflow automation, security tooling, Linux operations, and web products.
developer@github:~$ whoami
Ariful Anik
developer@github:~$ focus
AI workflow systems | security automation | Linux tooling | web products
developer@github:~$ current
Building Neurosformer and learning deeper AI agents, RAG, document intelligence,
human-in-the-loop workflows, and automation systems that turn messy operations
into structured, reviewable, auditable software.
I like engineering that comes from real friction: document chaos, recon noise, fragile systems, restricted networks, repetitive operations, and product workflows that need clearer structure. The goal is not to collect shiny demos. The goal is to build useful systems that survive contact with real users and real constraints.
I am building toward AI workflow intelligence for document-heavy and compliance-heavy industries: OCR, extraction, RAG with citations, human review, audit logs, checklist automation, and vertical products like NeuroDocOps, NeuroFashionOps, NeuroClinic Docs, and AgroNeuro.
Production recommendation pipelines for applicant, job, and event matching using retrieval signals, embeddings, graph signals, policy filters, telemetry, and baseline/LightGBM ranking.
I build recon helpers, subdomain workflows, cloud/SNI enumeration, payload notes, bug bounty utilities, and scripts that reduce repeated manual work.
I keep small operational tools around WSL maintenance, terminal setup, Raspberry Pi memory tuning, backup/config flows, notifications, and Linux automation.
Backend and APIs
Frontend and product UI
Data, storage, and product infrastructure
AI, RAG, recommendation, and research direction
Systems, security, and automation
Other things I have touched
principles/
+-- build tools that remove repeated work
+-- keep humans in the loop for risky AI workflows
+-- prefer citations, audit logs, and reviewable outputs
+-- make Linux do boring work automatically
+-- treat security as a default engineering habit
+-- ship small, learn fast, then harden what proves useful
