Shashank Bantupalli
Understand.. Don't Under Estimate
Intern · Engineering
Aspiring software developer with experience in backend development, machine learning, and data engineering. Proven ability to build end-to-end solutions and contribute to innovative projects.
Open to Work Intern Resume Verified Social Verified
Profile updated 5 days ago
Most Recent Role
Dev Intern
Helllo AI
May 2026 – July 2026
Education
Bachelor of Technology in Computer Science
Vellore Institute of Technology
GPA 8.51
Skills & Technologies
25 detectedLanguages
SSQL
Frameworks & Libraries
PPandas
NNumPy
Tools & Platforms
Other
LLightGBM
XXGBoost
YYOLOv8
OONNX
RREST APIs
JJWT Auth
JJetson Nano
LLiDAR SLAM
BBLDC Motor Control
EEmbedded C
Work Experience
H
Dev Intern
May 2026 – July 2026Helllo AI
- Built the backend for an AI-driven lead-enrichment platform end-to-end: designed the PostgreSQL schema and data pipelines that map and connect prospect identities across platforms.
- Built a cross-platform identity-resolution pipeline that takes a company's Instagram engagers and resolves each to their correct LinkedIn profile via public APIs.
- Shipped the product's frontend end-to-end, closing the loop from raw social ingestion to a working, demoable dashboard.
- Built a lead-scoring engine that ranks prospects by intent signals.
Education
Aug 2024 – Present
Vellore Institute of Technology
Bachelor of Technology in Computer Science
8.51
GPA
25
Skills
1
Experiences
Projects
3 built- Wrapped a production-grade ML ensemble (LightGBM + MLP + Bayesian Ridge across 5 leagues, novel League Strength Coefficient framework) predicting cross-league player performance translation into a full-stack product.
- Built a FastAPI backend with JWT authentication serving model predictions; migrating persistence to PostgreSQL and building a React dashboard with SHAP-based per-prediction explainability.
- Authored a formal SRS covering 25 functional and 15 non-functional requirements, UML diagrams, and a risk matrix ahead of CI/CD rollout.
- Authored a research paper benchmarking RAG defenses (EcoSafeRAG, ATM, Guardrails) against prompt injection, retrieval poisoning, and opinion manipulation across 3 SLA deployment profiles (SaaS, VPC, Regulated).
- Engineered a fully containerised microservices evaluation harness (Gateway → Hybrid Retriever → pgvector → Policy Middleware) with pinned model commits for deterministic reproducibility; open-sourced on GitHub
- Predicting battery SoH and Remaining Useful Life using an AttentiveLSTM model.
- Built an AttentiveLSTM predicting battery SoH and Remaining Useful Life (RUL) with a full ONNX/TFLite export pipeline for a 32 TOPS edge NPU; attention interpretability analysis for safety-critical deployment.
Powered by DevsUnite AI
AI résumé parsing · match scoring · cover letters · interview prep