Sanitized resume
Arham Hassan
AI systems builder. Queen's Computer Science (AI). Developer intern at Accreva. Founder of Elvyna.
Profile
Third-year Computer Science (AI) student at Queen's. Developer intern at Accreva supporting product development, currently with a frontend-heavy focus. Founder of Elvyna. Hands-on proof across AI governance (Halluci-NOT), citation-grounded RAG (Zahan Ed), AI automation, voice agents, and real-time robotics (VEX U). SAP Generative AI Developer certified.
Selected Projects
Accreva Business Ledger
Case studyDeveloper intern work across the Accreva product cycle: frontend app development, integration support, FastAPI domain API, MCP tool layer, and ledger-safe agent workflows.
- Supporting the lead developer across the repo, with current work focused mainly on frontend app development.
- Phase C active: partners, payment terms, invoices, and payment workflows in code.
- API-only write path enforced. No model output touches Postgres directly.
Elvyna
Case studyAI receptionist for trades: Retell handles voice intake, Twilio captures missed-call/SMS handoff, n8n orchestrates the workflow, OpenAI structures the lead, and the owner gets a text summary in under 10 seconds.
- Offer designed around a specific missed-call pain trades businesses already feel.
- Retell agent built for flexible voice intake across messy caller requests.
- End-to-end workflow mapped: missed call to AI intake to owner brief to follow-up.
Zahan Ed
Case studyOffline RAG pipeline: textbook → chunk with page metadata → Sentence Transformers embeddings → FAISS retrieval → citation-enforced generation → hard refusal on out-of-scope queries. Runs on CPU. LAN-deployable. 4-8GB RAM.
- 85% citation accuracy on a structured evaluation set.
- 78% page accuracy. The system knows where the answer came from.
- 100% correct refusal on out-of-scope prompts.
Halluci-NOT
Case studyTwo-layer governance pipeline: regex and allowlists catch structural violations deterministically, GPT-4o mini handles context-sensitive reasoning. Every output scores 0-100 and routes to PASS, REVIEW, or BLOCK.
- PII exposure blocked at severity 95.
- Unapproved vendor blocked at severity 85.
- Fake SAP T-code ME99X blocked at severity 80.
Technical Skills
Languages
Python, TypeScript, Java, C/C++, SQL, Rust
AI systems
RAG, embeddings, prompt engineering, function calling, OpenAI, SAP AI Core, agent orchestration, MCP, Retell
Backend
FastAPI, Postgres, SQLAlchemy, audit-first service design
Frontend and ops
React, Next.js, Tailwind, Git, Docker, n8n, Twilio, Retell
Robotics
Rust, Python, OpenCV, sensor fusion, real-time object detection
Certifications
SAP Generative AI Developer
SAP Learning, January 2026. Enterprise AI systems design, SAP AI Core, and GenAI Hub integration. Rare credential at third-year undergraduate level.
Education
Queen's University. B.Comp (Honours), Computer Science with AI focus.
Private academic details and private contact details intentionally excluded.