Sankalp Mallappa

AI Engineer · Machine Learning · Computer Vision · NLP
Hyderabad, India · mallappasankalp@gmail.com
LinkedIn • GitHub

About Me

I’m an AI Engineer with ~1 year of hands-on experience deploying real-world ML solutions.
I've worked across medical imaging, cloud automation, and e-commerce, leveraging deep learning, NLP, and LLMs.
I'm seeking full-time Machine Learning or Data Science roles starting August 2024.

Education

B.Tech in Electronics and Instrumentation Engineering
Minor in Data Science
Manipal Institute of Technology, Manipal
Expected Graduation: 2024     |    CGPA: 7.64

Intermediate (MPC)
Delta Junior College · 2017–2019 · 84.9%

High School
Delta High School of Science

Skills

Programming & Tools:
Python · PyTorch · Scikit-learn · Pandas · NumPy · MLflow · Matplotlib
Transformers · LangChain · AWS · FAISS · Streamlit

Soft Skills:
Problem Solving · Communication · Collaboration · Learning Agility

Experience

AI Developer – FinOps Automation

Orange People (Remote) · Apr 2025 – May 2025
- Built an LLM-based interface to automate AWS services (EC2, RDS, S3) via Boto3.
- Designed prompt pipelines and JSON parsers for intent detection and parameter extraction.

AI Engineer – Freelancing

Remote · Nov 2024 – Mar 2025
- Developed CV pipelines to extract metrics from microscopic jute fibre images.
- Built ultrasound models to measure fetal head circumference, improving clinical workflow accuracy.

Machine Learning Intern

Akridata, Bengaluru · Jan 2024 – Sept 2024
- Refined OWL-ViT prompts for object detection and autolabeling in Data Explorer.
- Built a chatbot to query pandas DataFrames using natural language and chain-of-thought prompting.
- Created lightweight tracking tools for video datasets.

Deep Learning Intern

INCOIS, Hyderabad · Dec 2022 – Jan 2023
- Built a YOLOv7-based instance segmentation model to detect fishing zones via satellite imagery.
- Achieved 94% IoU; delivered a proof-of-concept for ocean resource monitoring.

Projects

| Project | Description | Link | |--------|-------------|------| | Multi-Document RAG | Streamlit app for querying PDFs using FLAN-T5 and FAISS | GitHub | | OCT Image Classification | CNN + SVM classifier for 100k+ retinal OCT images (93% acc) | — | | House Price Prediction | Regression model with engineered features, +2.5% gain | — | | Line Follower Bot | Built with ATmega16 and IR sensors (video linked) | Watch Demo |

Publications

• YOLOv5s-CAM: A Deep Learning Model for Automated Detection and Classification of Intracranial Hematoma in CT Images
  IEEE Access, Dec 2023
  Enhanced YOLOv5s with CAM for improved hematoma detection and classification.

• Efficient Detection of Intracranial Hematoma Using Window-Based Stacking and YOLOv5
  ICAECA 2023
  Proposed a YOLOv5-based pipeline with window-stacking ensemble to detect hematomas from CT scans.