Fullstack Engineer with less than a year in Web Applications & AI/ML
AI is analyzing your overall score…
Identifying your key strengths…
Evaluating your skill match against the job requirements…
Assessing your cultural and operational fit
Results-driven Software Engineering undergraduate at FAST-NUCES (7th semester) with a proven track record of shipping production-grade, enterprise-ready full-stack applications, automated QA suites, and database-driven systems. Delivered a hackathon-winning health-tech product competing against multi-university teams. Maintains 90%+ unit test coverage, enforces security-first development (bcrypt, PDO, session auth), and builds scalable REST APIs that decouple frontend and backend for independent growth. Seeking opportunities to contribute to impactful engineering teams as a software developer, software architect, or QA engineer.
FAST National University (FAST-NUCES)
BS Software Engineering
August 1, 2023 – June 30, 2027
DIY Recipe Builder - Full-Stack Web Application
April 1, 2026 – April 1, 2026
Shipped production full-stack recipe platform following industry-standard development pipelines with 12+ REST API endpoints; eliminated SQL injection attack surface entirely through bcrypt hashing, session auth, and PDO prepared statements. Engineered interactive drag-and-drop ingredient UI with real-time calorie tracking using vanilla JavaScript, backed by a fully normalized 3NF MySQL schema across 8 tables. Achieved sub-200ms average API response time by indexing hot query paths and caching ingredient lookups — improving perceived performance by ~40% over initial baseline.
Fake News Detection - Research Replication & Improvisation (ISOT Dataset)
April 1, 2026 – April 1, 2026
Replicated a published ML benchmark on the ISOT Fake News Dataset (44,919 articles), reproducing SVM F1 = 0.9967 using TF-IDF unigrams with 3-fold stratified cross-validation — confirming paper results within ±0.001. Surpassed the baseline through 5 principled improvements: extended TF-IDF to word bigrams (ngram 1,2), added character n-gram features (char_wb, 3-5), fused both sparse matrices via scipy.sparse.hstack, and tuned SVM/PA hyperparameters via GridSearchCV — pushing best F1 to 0.9976 (+0.09%). Conducted a rigorous 5-configuration ablation study isolating each improvement's individual contribution, and built a Voting Ensemble (SVM + Passive Aggressive + Logistic Regression) with CalibratedClassifierCV achieving F1 = 0.9975 with improved stability. Matched and slightly exceeded the original paper's best classical ML result (SVM F1 = 0.9976 vs paper ~0.99) without any deep learning or external data — all improvements grounded in course-covered techniques.
Mindernity - Postpartum Mental Health Platform
January 1, 2026 – June 1, 2026
Identified critical market gap: 26.9% postpartum depression prevalence (AKU) across 6M+ annual Pakistani births, with zero affordable Urdu-language mental health tools available — translating insight into a tri-module solution for mothers, husbands, and family members. Architected product featuring an Urdu-first AI chatbot, EPDS-validated weekly mood check-ins, anonymous peer groups, and Rs. 300/session teletherapy — undercutting existing private options by over 90%. Modeled unit economics achieving break-even at 1,500 subscribers (Month 6-8) on a Rs. 299/month subscription; projected Rs. 3L/month burn rate with Year-1 reach of 10,000+ mothers and 30,000+ family members. Designed a 4-phase market expansion: Rawalpindi/Islamabad OPDs → major city NGOs → secondary cities via CHWs → rural Pakistan via USSD/voice — targeting 200K+ users.
Database-Driven Game Store - E-Commerce Backend
May 1, 2025 – November 1, 2025
Designed enterprise-grade e-commerce backend across 10 normalized tables with full referential integrity and indexed queries — reducing average query response time by 35% compared to un-indexed baseline. Built Node.js RESTful API layer to fully decouple frontend and backend, enabling independent scalability; documented 20+ endpoints enabling faster onboarding for future contributors.
Automated QA Suite — Web Application Testing
January 1, 2025 – December 31, 2025
Attained 92% code coverage across 15 test classes using JUnit and Mockito — exceeding the 70–80% industry benchmark by over 15 percentage points. Automated cross-browser UI validation across Chrome and Firefox for 3+ web apps using Selenium WebDriver, eliminating ~6 hours/week of manual regression effort.
Emotion Detection System — ML Pipeline
January 1, 2025 – December 31, 2025
Trained end-to-end emotion classification model achieving 84% accuracy across 6 emotion categories; handled full preprocessing pipeline and feature extraction via OpenCV and supervised ML algorithms.
Hybrid Semantic Retrieval & Intelligence System (HSRIS)
January 1, 2025 – December 31, 2025
Built a multi-stage NLP pipeline from scratch using base PyTorch to process and retrieve 8,470+ customer support tickets — implementing Label Encoding, One-Hot Encoding, TF-IDF, and GloVe embeddings without any high-level wrappers. Engineered a hybrid retrieval engine combining TF-IDF keyword search and GloVe semantic embeddings with a weighted FinalScore = α(TF-IDF) + (1-α)(GloVe), returning top-5 relevant tickets with Precision@5 evaluation. Optimized performance using torch.sparse tensors to prevent RAM overflow and torch.nn.DataParallel across dual T4 GPUs, parallelizing similarity calculations over batches of 100 queries. Deployed a Gradio app with an α-slider enabling real-time switching between keyword and semantic retrieval modes, displaying predicted Ticket Type and top-3 past resolutions.
Urban Environmental Intelligence Engine — Air Quality Analytics
January 1, 2025 – December 31, 2025
Architected a modular Python pipeline (no Jupyter) ingesting hourly PM2.5, PM10, NO2, Ozone, Temperature, and Humidity data from 100 global sensor nodes via the OpenAQ API across all of 2025. Applied PCA-based dimensionality reduction on 6 environmental variables to produce 2D cluster visualizations distinguishing Industrial vs. Residential zones, with loading analysis identifying primary pollution drivers. Designed high-density temporal visualizations to surface periodic pollution signatures across 100 simultaneous time-series, identifying 24-hour traffic cycles vs. 30-day seasonal shifts without overplotting. Deployed an interactive Streamlit dashboard; applied Lie Factor and Data-Ink Ratio principles to reject misleading 3D charts in favor of bivariate mapping with perceptually accurate sequential color scales.
NYC Congestion Pricing Audit - Big Data Transportation Analysis
January 1, 2025 – June 1, 2026
Built an automated ETL pipeline ingesting 100M+ row NYC TLC Yellow & Green taxi parquet files for all of 2025 using Big Data tooling (no pd.read_csv on full dataset), unifying schemas and performing all aggregations before converting to Pandas. Implemented a Ghost Trip fraud detection filter flagging impossible physics (speed > 65 MPH), teleporter trips (< 1 min, fare > $20), and stationary rides — logging flagged records as an Audit Log for top-5 suspicious vendors. Conducted geospatial surcharge compliance audit for Manhattan Congestion Zone (south of 60th St), identifying top-3 pickup locations with highest rates of missing surcharges and quantifying Q1 2024 vs. Q1 2025 trip volume decline. Modeled Rain Elasticity of Demand by joining NOAA precipitation data with daily trip counts; delivered a 4-tab Streamlit dashboard covering border effect choropleth, velocity heatmaps, tip crowding-out analysis, and weather elasticity scatter plots.
UI/UX Prototypes & OOP Systems Projects
January 1, 2023 – December 31, 2025
Delivered high-fidelity Figma prototypes for 2 full systems applying UX research, WCAG accessibility standards, and visual hierarchy across patient dashboards and booking flows. Completed 5+ progressive academic projects across 4 semesters covering OOP, DSA, OS concepts, file I/O, and real-time logic — including an Arabic Dictionary App in Java Swing/AWT.
Cultural Fit Analysis
The candidate's diverse project portfolio, ranging from full-stack web development to advanced machine learning and big data analytics, indicates a strong curiosity and willingness to explore various technical domains. Their involvement in a health-tech platform addressing a critical social issue (postpartum mental health) and leadership roles in university societies suggest a proactive and community-oriented mindset. This breadth of interest and drive to create impactful solutions aligns well with a culture that values innovation, continuous learning, and social responsibility. However, the lack of professional experience means their adaptability to corporate cultural norms and long-term team collaboration dynamics is an area for further assessment.
Soft Skills & Operational Fit
The candidate demonstrates strong initiative and a results-driven mindset, as evidenced by leading a hackathon-winning project and consistently aiming for high performance metrics (e.g., 92% code coverage, sub-200ms API response times). Their project descriptions highlight a structured approach to problem-solving and an understanding of real-world constraints (e.g., market gaps, unit economics). The leadership roles in university activities suggest good organizational and teamwork skills. However, without direct work experience, the ability to navigate complex team dynamics and operational challenges in a professional setting is yet to be fully validated.