Jaisha Shahid

CS at LUMS | Agentic AI & HCI Researcher | Building Human-Centered Intelligent Systems

Overview

Computer Science student at LUMS specializing in the intersection of Agentic AI and Human-Computer Interaction. Currently engineering multi-agent workflows at Appsgenii, targeting 85% automation through reliability-focused monitoring. Experienced in leading large-scale user studies (40+ participants) and translating complex accessibility needs into iterative system designs. A rare blend of high-scale engineering and empathetic research.

Pakistan

Experience

WORK

AI Agentic Systems Engineering Intern

Appsgenii Technologies

Assisted in building and testing LLM-powered multi-agent workflows using frameworks such as LangGraph/OpenAI APIs to automate customer interactions, triage tasks, and internal knowledge queries. Contributed to the design and evaluation of production-grade agentic AI systems for enterprise workflows, including service automation and knowledge-retrieval pipelines (RAG-based architectures). Worked with engineering teams on optimizing AI system accuracy, latency, and reliability in real-world deployments handling large-scale enterprise usage.

•Assisted in building and testing LLM-powered multi-agent workflows using frameworks such as LangGraph/OpenAI APIs to automate customer interactions, triage tasks, and internal knowledge queries

•Contributed to the design and evaluation of production-grade agentic AI systems for enterprise workflows, including service automation and knowledge-retrieval pipelines (RAG-based architectures)

•Worked with engineering teams on optimizing AI system accuracy, latency, and reliability in real-world deployments handling large-scale enterprise usage.

Key Achievements

→Developing an agent-to-agent monitoring framework using confidence ratios and behavioral triggers to automate quality control.

→Implementing caching strategies and optimizing retrieval logic to minimize redundant database calls and reduce system latency.

→Targeting 85% automation of quality control processes through agent-to-agent monitoring.

→Optimizing retrieval logic and caching to minimize redundant database calls.

LLMLangGraphOpenAI APIsRAGAgentic AI

2025 – 2026

WORK

Teaching Assistant

Lahore University of Management Sciences

Served as TA for two consecutive sections of the same course under the same instructor, supporting 150+ students per semester across both sections. Facilitated student-professor communication, graded 300+ quizzes and assignments per semester, and maintained accurate course records throughout the term.

•Served as TA for two consecutive sections of the same course under the same instructor, supporting 150+ students per semester across both sections

•Facilitated student-professor communication, graded 300+ quizzes and assignments per semester, and maintained accurate course records throughout the term

Key Achievements

→Supported 150+ students per semester

→Graded 300+ quizzes and assignments per semester

2025 – 2025

WORK

Student Partner

Peer Pedagogical Partnership, LUMS

Collaborated with Faculty Partner to design AI integration for the Data Analytics for New Product Development course, reaching 60+ students. Created 12+ AI-centered tools, chatbots, course materials, and curriculum for Data Analytics (R/Python). Optimized the flagship coding-integrated marketing course for future iterations by systematically recording and analyzing student feedback.

•Collaborated with Faculty Partner to design AI integration for the Data Analytics for New Product Development course, reaching 60+ students

•Created 12+ AI-centered tools, chatbots, course materials, and curriculum for Data Analytics (R/Python)

•Optimized the flagship coding-integrated marketing course for future iterations by systematically recording and analyzing student feedback

Key Achievements

→Designed and integrated 12+ AI-centered tools and chatbots for Data Analytics curriculum, directly impacting 60+ students.

→Developed an R language lab assistant chatbot to guide students through coding logic and product development workflows, reducing manual troubleshooting during lab sessions.

→Systematically analyzed student feedback to optimize coding-integrated marketing courses for future iterations.

→Reached 60+ students

→Created 12+ AI-centered tools, chatbots, course materials, and curriculum

AIRPython

2024 – 2024

INTERNSHIP

Marketing Intern

U.N.I. Social Networking App

Led marketing campaigns across 3 platforms as content writer, videographer, graphic designer, market researcher, and social media handler. Grew Instagram engagement and views by 40%; produced 10+ videos each surpassing 1 million views.

•Led marketing campaigns across 3 platforms as content writer, videographer, graphic designer, market researcher, and social media handler

•Grew Instagram engagement and views by 40%

•Produced 10+ videos each surpassing 1 million views

Key Achievements

→Grew Instagram engagement and views by 40%

→Produced 10+ videos each surpassing 1 million views

Instagram

Project Portfolio

PERSONAL

Present

Real-Time Multiplayer Ludo Game

Developer

Developed a real-time multiplayer game with lobbies, live state updates, chat, turn-based logic, authentication, history, and leaderboard APIs using the MERN stack; supporting 4 concurrent players per game, and an unlimited number of rooms.

ReactTypeScriptExpress.jsSocket.IOMongoDB AtlasMERN stack

PERSONAL

Present

Asynchronous Email Notification System

Developer

Developed an automated monitoring tool to eliminate manual database checks by triggering real-time email alerts on PostgreSQL insert events, ensuring immediate visibility into data changes without direct DB access.

SupabasePostgreSQLpg_netResendServerless Backend

PERSONAL

Present

Azure Cloud-Native Application Architecture

Cloud Architect

Architected a secure, private-network framework to streamline service communication and automate report generation, ensuring data storage and processing remained entirely off the public internet.

Azure App ServiceFunctionsACRVNetsPrivate EndpointsDNS

PERSONAL

Present

Research on AI Smartphone Guidance Sandbox for Older Adults

Developer

Designed and tested an accessibility prototype with AI voice chat integration to support older adults in completing smartphone tasks. Led an iterative design process, evolving from static highlighted buttons to animated finger pointers and finally synchronized moving indicators with verbal instructions after discovering that static cues were insufficient for the target demographic.

AndroidJavaKotlinXMLView BindingGradleGit

Education

2023 – 2027

BSc Computer Science

Lahore University of Management SciencesSpecialization in Computer Science

GPA: 3.347

Societies & Activities

LUMS Women in ComputingLUMS Mathematics Society

Impact & Recognition

Awards

A Levels Institution

100% Merit Scholarship – A Levels

LUMS Mathematics Society

Pi Day Winner

A Levels and O Levels Institution

Honor Roll – A Levels and O Levels

Social Impact & Activities

2025 – 2026

LUMS Women in Computing

Director Events

Planned and executed 7 national events attended by 3,000+ participants in total, managing end-to-end logistics and coordination. Secured and managed Rs. 700,000 in sponsorships across 2 corporate partners. Led outreach, event strategy, and community-building initiatives promoting women's participation in computing through recruitment campaigns, technical workshops, and peer engagement.

EngagementORGANIZER