AI Researcher • Engineer • Patent Holder

Spriha
Podder

AI should not just predict. It should understand.

Bridging perception, reasoning, and emotional intelligence in intelligent systems.

About

Building intelligence
that matters.

I am an AI researcher and engineer focused on building intelligent systems that are accurate, interpretable, and deployable. My work spans computer vision, natural language processing, explainable AI, and agentic systems.

Indian Patent holder. German Research Grant recipient. Published at ASCE 2025 in Arizona. UCL London Summer School alumna. Exchange scholar at ESIGELEC, France. Interned at Deloitte India. I bridge research and production, transforming ideas into working AI systems.

Education

B.Tech Computer Science — Manipal University Jaipur

Expected 2027 • CGPA 9.84/10 • Dean's List

Global Exposure

UCL London • ESIGELEC France

Data-Driven Web Apps & AI • Automotive Embedded Systems

Focus Areas

CV • NLP • XAI • GenAI • Agentic AI • MLOps

9.84

CGPA / 10

4x

Dean's List

2

Patents / Grants

4

Research Papers

Recent Interests · Articles

Thinking out loud
about where AI is actually going.

Short, opinionated pieces on the frontier I'm obsessed with — ethics, interpretability, healthcare XAI, and the new ML edge in quant. Each one ships with an interactive visualization.

CapabilityAlignmentTransparencySafetyGame-Theory Stability

Ethics Compass

Game-theoretic governance vs. laissez-faire vs. static guardrails.

AI Ethics · Governance6 min

Claude, Fable & Mythos — why ethical AI is the only real moat left

Anthropic's Claude refused to ghostwrite a viral fanfic. Mythos got jailbroken into impersonating a real teen. Fable shipped a celebrity deepfake before lunch. The same week. Same news cycle. Same wake-up call: alignment is not a feature, it's the product.

Read & interact

Sparse Features

SAE-decoded monosemantic clusters: syntax · entities · math · code · deception.

Interpretability · Transformers7 min

Transformers are not black boxes — sparse autoencoders cracked the lid

LLMs and RAGs both stand on one backbone: the transformer. We treat it as a black box because the residual stream is dense and polysemantic. Anthropic's sparse autoencoder work showed those tangled features actually live on a clean geometric manifold — surprisingly close to how humans organize meaning.

Read & interact

XAI Backtrace

Prediction → layer → attention → Grad-CAM → clinician trust.

Explainable AI · Clinical5 min

Explainable AI in healthcare — why backtracing is non-negotiable for clinical trust

94% accuracy doesn't move a single clinician if they can't see why. XAI is the chain of custody between a model prediction and a radiologist's pen. Grad-CAM, LIME, SHAP and integrated gradients aren't cosmetics — they're the audit trail medicine has demanded since the Hippocratic oath.

Read & interact
201820192020202120222023202420252026100%50%0%Classical factor alphaDL-augmented signals

Factor crowding has crushed classical alpha. Transformer-based regime detection, diffusion priors on order flow, and offline RL execution policies are the new edge — each compounding while the old edge melts.

DL Frontiers · Quant Finance6 min

The new edge — how transformer & diffusion advances are eating quant

Classical factor alpha has been crowded into oblivion. The 2026 edge is built from offline RL execution, transformer-based regime detection, diffusion priors over order-flow, and SAE-interpretable feature stores. The quants who treat ML as a black box are about to get arbitraged by the ones who treat it as a microscope.

Read & interact

Experience

Where I've
made impact.

Deloitte

Technology & Transformation Intern

Jun 2025 — Jul 2025

Delhi, India

  • Analyzed client business processes and technology workflows to support enterprise transformation.
  • Researched cloud and data technologies including Azure Databricks, Apache Spark, and ERP integration.
  • Contributed to development and testing of scalable, user-centric solutions.

LEA Associates South Asia

Technical Apprentice

May 2021 — Jun 2021

Delhi, India

  • Supported predictive modeling for road maintenance using pavement defect analytics.
  • Built ANN-based spreadsheet models to identify major road construction cost drivers.
  • Managed database administration and contributed to website design initiatives.

Research & Patents

Published. Patented.
Peer-reviewed.

Indian Patent — Jul 2025German Research Grant — Oct 2025

A Lightweight Multi-Task Deep Learning Framework for Simultaneous Classification of Alzheimer's Disease Stages and Parkinson's Disease with Explainable AI

  • Multi-task CNN with two classification heads and modified combined loss function
  • ResNet18-based architecture integrated with clinical data for improved accuracy
  • LIME, Grad-CAM, and SHAP for explainability — enabling clinical trust
  • Unified, interpretable diagnosis of neurodegenerative diseases from brain MRIs

ASCE International Conference on Transportation & Development — Glendale, Arizona, USA • Jun 2025

View all on ResearchGate

Diffusion-based Model for Price Adjustment Dynamics in Delayed Road Construction

Lead modeling and simulation. Implemented RBF networks for time-series price diffusion.

Cost Variation Prediction: Fuzzy Logic and Clustering Approach

Co-designed fuzzy inference system. Applied K-means clustering for budget forecasting.

RBF Regression and Bootstrapping for Price Adjustment in Road Contracts

Built regression framework with bootstrapped sampling for variance-aware predictions.

Hazard-Based Model for Financial Milestones in Road Construction

Survival analysis-based hazard model for time-to-milestone risk prediction.

Projects

Shipped. Competed.
Recognized.

Tap any project to dive into architecture, results, images, demo video and the GitHub repo.

Computer Vision Portfolio

Seeing beyond pixels.
Understanding through vision.

From medical imaging to real-time detection — building vision systems that are accurate, interpretable, and deployment-ready.

Featured Research

Medical Imaging / XAI

Multi-Task Neurodegenerative Disease Detection

Alzheimer's & Parkinson's Classification from Brain MRIs

Overview

A lightweight multi-task deep learning framework that simultaneously classifies Alzheimer's disease stages and Parkinson's disease from brain MRI scans, integrated with Explainable AI for clinical trust and transparency.

Problem

Current diagnostic pipelines treat Alzheimer's and Parkinson's as separate classification tasks, requiring multiple models, redundant computation, and inconsistent interpretability — limiting real-world clinical deployment.

Technical Approach

Designed a shared-backbone multi-task CNN with two parallel classification heads
Modified combined loss function balancing both tasks for joint optimization
ResNet18-based feature extraction integrated with tabular clinical metadata
Implemented LIME, Grad-CAM, and SHAP for multi-layered explainability
End-to-end pipeline from MRI preprocessing to interpretable diagnosis

Architecture Pipeline

Input: Brain MRI Scans + Clinical Data
Backbone: ResNet18 (shared feature extractor)
Head 1: Alzheimer's Stage Classification (4 classes)
Head 2: Parkinson's Binary Detection
XAI Layer: Grad-CAM heatmaps + LIME regions + SHAP values
Output: Joint diagnosis with visual explanations

Key Results

94.2%

Alzheimer's Accuracy

96.8%

Parkinson's Accuracy

~40%

Model Size Reduction

<200ms

Inference Time

Techniques & Tools

Multi-Task LearningResNet18Grad-CAMLIMESHAPTransfer LearningMRI PreprocessingClinical Data Fusion

Impact

Indian Patent filed (Jul 2025). German Research Grant awarded (Oct 2025). Framework enables unified, interpretable neurodegenerative diagnosis — bridging the gap between deep learning accuracy and clinical trust.

Applied Vision Projects

From theory to real-time.

Hands-on implementations spanning gesture recognition, real-time tracking, augmented reality, and intelligent monitoring systems.

11Projects
C++PythonOpenCVMediaPipeYOLOv8
01Real-Time Color Segmentation

Red Pixel Detection

Real-time red color detection using webcam input and HSV color space segmentation. Dual HSV thresholds handle hue wrapping with morphological filtering for noise reduction. Displays original frame, mask, and segmented result simultaneously.

C++OpenCVHSV Color SpaceMorphological Ops
02Object Tracking

Red Pixel Centroid Tracking

Extended red color detection with image moments to compute the centroid of detected regions in real time. Tracks object motion via dynamic centroid overlays, contour filtering, and moment-based coordinate computation for stable tracking.

C++OpenCVImage MomentsContour Detection
03Interactive CV

Gesture-Controlled Balloon Pop Game

Interactive gesture-based balloon popping game using real-time hand tracking. MediaPipe detects fingertip landmarks mapped to a virtual game canvas with collision detection, animated sprites, scoring logic, and leaderboard storage.

PythonOpenCVMediaPipeNumPyCollision Detection
04Centroid-Based Drawing

Virtual Remote Painter

Real-time gesture-based drawing interface using centroid tracking of a colored object. Computes centroids via image moments, renders motion trails on a virtual canvas with color selection zones and interactive painting controls.

C++OpenCVHSV SegmentationImage MomentsReal-Time Graphics
05Surveillance System

Motion Detection & Intrusion Alarm

Adaptive background modeling system using running average for motion detection. Gaussian blur and morphological filtering minimize noise. Triggers asynchronous alarms and auto-saves timestamped snapshots during intrusion events.

C++OpenCVBackground SubtractionContour DetectionAsync Audio
06AR Background Substitution

Invisibility Cloak

Augmented reality illusion detecting a specific cloak color via HSV segmentation. Generates masks to isolate cloak regions and replaces them with a previously captured background frame with morphological smoothing for seamless blending.

C++OpenCVHSV DetectionMaskingImage Compositing
07Touchless Musical Interface

Gesture-Controlled Virtual Piano

Touchless virtual piano controlled entirely by fingertip gestures. MediaPipe hand tracking detects finger positions mapped to key regions as sound triggers with cooldown constraints and threaded audio playback.

PythonOpenCVMediaPipePygameThreaded Execution
08Touchless Navigation

Gesture-Controlled Dataset Viewer

Touchless dataset explorer using swipe and pinch gestures. Wrist displacement tracking for navigation and Euclidean distance measurement for zoom control with dynamic image resizing and highlighted thumbnails.

PythonOpenCVMediaPipeNumPyGesture Recognition
09Gesture-Based Input

Virtual Mouse & Game Controller

Vision-based virtual mouse where fingertip movements control the cursor. Pinch detection for click/drag operations with smoothing filters to reduce jitter, mapped from webcam coordinates to full-screen resolution.

PythonOpenCVMediaPipePyAutoGUISmoothing Filters
10Image + Webcam Detection

YOLO Object Detection

YOLO-based object detection for static images and live webcam streams. YOLOv8 processes frames for bounding box prediction and classification with labeled real-time outputs for surveillance, automation, and smart monitoring.

PythonOpenCVYOLOv8 (Ultralytics)NumPy
11YOLO + Sustainability Modeling

Real-Time Ecological Impact Indicator

Real-time environmental monitoring combining YOLO object detection with lifecycle emission modeling. Assigns ecological weights using DEFRA/EPA/IPCC datasets, computes per-object impact with logarithmic normalization, and aggregates into a scene-level Eco Index (0–100).

PythonOpenCVYOLOv8PandasLifecycle Assessment

“Vision is not just seeing — it's understanding what you see.”

Skills & Recognition

The toolkit.
The track record.

Languages

PythonCC++JavaScript

AI / ML

MLDLNLPCVXAIGenAIAgentic AILLMsRAGsGame TheoryStatisticsInferencing

Web & Backend

React.jsNode.jsFastAPIFlaskStreamlitTailwind

Tools & Infra

GitBashLinuxMLOpsSQLROS2Embedded CEmbedded Linux

Recognition & Awards

4x Dean's List — Manipal University Jaipur
2x Student Excellence Award — MUJ
National Semi-Finalist — Flipkart Grid 7.0
Top 10 — HackX 3.0 International Hackathon
Round 2 — EY Techathon 6.0 (1,000+ teams)
UCL Summer School — First Division
Indian Patent Holder
German Research Grant Recipient
4x ASCE Published Researcher
CodeChef Bronze — 50+ DSA Problems

Contact

Let's build
something intelligent.

Open to AI/ML research collaborations, product-driven engineering roles, and opportunities that demand ownership and impact.

Delhi, India

SP.

Designed with precision. Built to last.