Help & Supportcareers@gravityer.com
AI Engineer with less than a year in AI/ML & Computer Vision
AI is analyzing your overall score…
Identifying your key strengths…
Evaluating your skill match against the job requirements…
Assessing your cultural and operational fit
Builds and ships end-to-end AI applications, not prototypes. Delivered four production-ready AI systems during a full-time internship at CodeCelix Technologies, integrating YOLOv8, Vosk, SAM, MiDaS, and faster-whisper into benchmarked Python and FastAPI pipelines. Engineers explicitly for offline-first, CPU-only environments: no cloud dependency, no GPU, sub-second latency targets, with every pipeline stage benchmarked independently before integration.
COMSATS University Islamabad
BS · Computer Science
August 1, 2022 – June 30, 2026
CodeCelix Technologies
AI Engineer Intern
December 1, 2025 – March 1, 2026
Rawalpindi, Punjab, Pakistan
AI Room Scan & Interior Styling System
January 1, 2023 – June 1, 2026
4-stage CPU-only pipeline: MiDaS depth estimation → YOLOv8n object detection → SAM instance masking → Stable Diffusion inpainting. Produces photorealistic redesigns from a single photo in under 2 min on a laptop CPU with no GPU. Applied INT8 quantization via ONNX Runtime on the detection stage; profiled each stage independently, identifying SAM mask generation as the primary bottleneck (~45s/image) for targeted optimization.
View ProjectReal-Time Speech Transcription Engine
January 1, 2023 – June 1, 2026
CPU-only real-time transcription tool using faster-whisper (Whisper-small) with WebRTC VAD-based audio segmentation. Achieved ~280ms transcription latency; tuned VAD thresholds to suppress false triggers from ambient noise and keyboard input.
View ProjectAR Platform 3D Model Viewer & QR System
January 1, 2023 – June 1, 2026
WebAR platform for viewing 3D product models in augmented reality via QR code scan no app install required. Python backend auto-generates QR codes linked to hosted GLB assets in Supabase Storage. Live at arviewer-seven.vercel.app.
View ProjectRAG Knowledge Base Management System
January 1, 2023 – June 1, 2026
Built a Retrieval-Augmented Generation pipeline for an AI voice agent team: document ingestion, semantic chunking, embedding generation, and vector-store retrieval (ChromaDB) exposed via FastAPI reducing hallucination rate vs. prompt-only baseline.
View ProjectGIDEON Offline Voice Automation Agent
January 1, 2023 – June 1, 2026
Fully offline, privacy-preserving desktop agent: Vosk handles local speech recognition (zero network calls), and a custom rule-based intent parser dispatches commands to OS-level automation functions. Achieved sub-500ms end-to-end response. Chose Vosk over cloud STT APIs to eliminate network dependency and protect sensitive user data a deliberate architectural tradeoff for offline-first, privacy-sensitive deployment environments.
View ProjectCultural Fit Analysis
The candidate's portfolio showcases a strong inclination towards building practical, deployable AI solutions, which aligns well with product-focused engineering cultures. The diversity of projects (computer vision, voice automation, RAG, AR) indicates adaptability and a broad interest in AI applications. The focus on optimizing for specific constraints (CPU-only, offline, privacy) suggests a resourceful and innovative mindset. The candidate's proactive approach to learning and applying various technologies (MiDaS, YOLOv8, SAM, Stable Diffusion, Vosk, faster-whisper, ChromaDB, FastAPI) demonstrates a growth mindset and a willingness to tackle complex challenges, making them a good cultural fit for dynamic and technically demanding environments.
Soft Skills & Operational Fit
The candidate's project descriptions highlight a methodical approach to problem-solving, including independent stage profiling and bottleneck identification. The emphasis on documentation and measured benchmarks suggests a detail-oriented and results-driven work attitude. The choice of offline-first solutions indicates a practical, user-centric mindset, and the ability to deliver multiple systems within a short internship period points to strong execution and time management skills. The deliberate architectural tradeoffs made for specific deployment environments demonstrate critical thinking and an understanding of real-world constraints.