Applied AI Engineer & Computer Science Student
Yeabsira
Tesfaye
Applied AI engineer and Computer Science student with 3+ years of professional software development experience working with local and international clients. Previously led web and product development projects across e-commerce, events, and business platforms, and now focus on building AI-powered applications, LLM agents, automation workflows, and developer tools. Skilled in Python, TypeScript, Next.js, React, PostgreSQL, and modern AI stacks, with a strong ability to translate business needs into practical, user-focused solutions.
Professional Experience
Software Developer — Chiropedic Mattress Ltd
Melbourne, Australia · Remote
- Led a team of 2 developers upgrading, maintaining and optimizing 6 e-commerce websites across Australian retail brands.
- Improved user experience and conversion rates through website optimization, analytics integration, and A/B testing.
- Enhanced operational reliability by streamlining payment systems, strengthening security, and coordinating remote delivery across time zones.
Senior Web Designer & Developer — Yeab Future
Addis Ababa, Ethiopia
- Led a team of 5 developers, managing end-to-end delivery from requirements gathering and design to deployment and maintenance.
- Built websites, e-commerce platforms, and business applications for clients including the U.S. Embassy in Ethiopia, Arifpay, Flawless Events, Tradenet USA, RND Ethiopia and Kilimanjaro Ventures.
- Collaborated directly with non-technical stakeholders to translate business needs into technical solutions while implementing analytics, payment systems, and security enhancements.
Web Designer & Developer — Enkopa Summit
Addis Ababa, Ethiopia
- Contributed to an event that hosted 1,200+ guests and attracted 6,500 website users with 25,000+ page views.
- Conducted user research, information architecture, wireframing, UI design, and full-stack development of the summit platform, including online registration, ticketing, payment processing, onsite scanning, program tracking, analytics integration, and security frameworks to support a seamless event experience and operational visibility.
- Worked closely with founders and the event team, providing technical support, reports, and post-event improvements.
Co-Founder & CEO — Gofere Travels
Addis Ababa, Ethiopia
- Led product direction, operations, development planning, and early business execution for a travel-booking startup focused on simplifying authentic trip discovery and booking in Ethiopia.
- Worked across product design, customer discovery, vendor coordination, technical implementation, and go-to-market planning, gaining hands-on experience in ownership, prioritization, and building software around real-world business workflows.
Projects
Team: Tesfaye, Y. M., Mathew, S., Nguyen, N. T., Yoselia, C | Prof. Song Dandan
- Built an AI-powered code evaluation platform that assesses student repositories against instructor-defined rubrics using LLM agents and sandboxed code execution.
- Developed end-to-end workflows for rubric generation, repository evaluation, feedback reporting, class management, and asynchronous grading pipelines.
- Awarded Grand Prize, 16th National E-commerce "Innovation, Creativity & Entrepreneurship" Challenge, 2026.
- Built a collaborative AI-assisted workspace that transforms architecture discussions into editable system-design canvases and implementation-ready specifications.
- Implemented AI-powered architecture generation, specification creation, and collaboration workflows using Trigger.dev, Liveblocks, and React Flow.
- Developed project management, access control, and persistence systems that help teams convert unstructured technical conversations into structured, reviewable implementation artifacts.
- Built an AI-powered voice interview platform that generates custom interviews, conducts real-time conversations through voice AI, and delivers structured, evidence-backed performance reports.
- Developed an end-to-end interview workflow combining conversational interview creation, real-time Vapi voice interactions, automated transcript analysis, and AI-driven grading with detailed feedback and scoring.
- Engineered a production-ready architecture using Next.js, Prisma, Trigger.dev, and Vercel Blob with secure server-side prompt handling, background processing, multi-language support, and protected user data.
- Built an open-source terminal coding agent with tool use, memory, session management, and interactive developer workflows.
- Designed agent orchestration capabilities including sub-agents, task delegation, MCP integrations, and context management for complex multi-step workflows.
- Implemented safety, reliability, and developer productivity features such as approvals, sandboxing, audit logging, runtime controls, and terminal-based command interfaces.
- Contributed to Synheart AI's open-source Python SDK and CLI for persona-driven, schema-validated LLM interactions with built-in safety controls and budget tracking.
- Developed infrastructure for controlled AI execution, including persona contracts, structured JSON outputs, safety enforcement, token constraints, and cloud/local deployment support.
PathoFusion — Histopathological Image Classification for Breast Cancer Diagnosis
Team: Tesfaye, Y. M., Prof. Xiangfu Zhang
- Designed and evaluated a deep learning pipeline for axillary lymph node metastasis classification in breast cancer diagnosis.
- Conducted model evaluation and experimentation under academic supervision.
- Awarded Third Prize, Qingju AI "Artificial Intelligence+" Special Competition, July 2025.
Core Skills
Education
BSc, Computer Science and Technology
Beijing Institute of Technology
- Beijing Municipal Government Scholarship, 2024–present.
- Relevant coursework: Introduction to Artificial Intelligence; Brain Science & Brain-Inspired Intelligence.
- Focus areas: applied AI systems, LLM agents, software engineering, continual learning, personalization, and AI safety.
Summer School for Generative AI
Tsinghua University
- Relevant coursework and lectures: Jie Tang — The Road to AGI; Zhiyuan Liu — LLM Agents; Yuxiao Dong — LLMs and Multimodal Models; Wenguang Chen — Distributed Systems for Large Models.
Research Experience
Research Intern — Synheart AI
- Co-authored the Human State Interface (HSI) framework, defining Human State Vectors for representing physiological, cognitive, and behavioral state across AI systems.
- Built emotion and stress inference pipelines using multimodal physiological datasets and contributed to affective AI model development and evaluation.
Publications
How you say it matters: Personalizing LLM responses via dual time-scale closed-loop adaptation
Human state interface infrastructure
AM-SDS Nexus 7 — Autonomous Martian seed deployment system
Selected Client Portfolio
Delivered or contributed to web/software projects for organizations and brands including: