Experience
Vehicle Classification and Detection using Neural Networks
Detecting and classifying vehicles in images and video for traffic monitoring
- A YOLO + Faster R-CNN model built for real-world traffic and autonomous-vehicle scenarios. Awarded Best Paper Award at St Joseph Engineering College, March 2025.
What it does
- The model takes images or video frames and both localizes and classifies vehicles by type, making it usable for downstream applications such as traffic-flow monitoring or perception in autonomous-vehicle systems.
What I did
- Combined YOLO (for fast detection) with Faster R-CNN (for higher-precision classification), rather than relying on a single architecture, trading off speed and accuracy where each mattered most.
- Achieved 90% mean Average Precision (mAP), evaluated using IoU-based metrics rather than just raw accuracy, to properly reflect localization quality.
- Handled the full data processing and evaluation pipeline in Python using TensorFlow and OpenCV.
- The accompanying paper was awarded the Best Paper Award by the Department of CSE at St Joseph Engineering College.
Tech stack: Python, TensorFlow, OpenCV, YOLO, Faster R-CNN
AI Resume Scoring Agent
An LLM agent that scores resumes against a job description with forced structured output
- A self-directed project built specifically to gain hands-on experience with tool-calling, prompt engineering, and workflow automation – not just to claim familiarity with them.
What it does
- Given a job description and a resume, the agent returns a structured JSON object (not a paragraph to interpret) containing an overall score, sub-scores for skills match and experience relevance, a list of specific named gaps, and a one-line summary – ready to sort, filter, or feed into another system automatically.
What I did
- Used the LLM's tool-calling API to force structured JSON output via a declared schema, rather than parsing free text with regex, guaranteeing the response always matches the expected format.
- Applied deliberate prompt engineering: explicit scoring criteria per field, a few-shot example of the expected reasoning style, and ordering the schema so the model's reasoning field comes before the numeric scores, so reasoning conditions the score instead of just justifying it after the fact.
- Built a Streamlit UI for uploading resumes and viewing a ranked, sortable results table.
- Wrapped the scoring logic in a Flask API and connected it to an n8n workflow: a form submission triggers the agent via webhook, which then sends an email notification with the result, tested against a real SMTP sandbox (Mailtrap).
- Built a small Lovable app that posts structured data to an external webhook, to get direct hands-on experience with a no-code builder rather than secondhand familiarity.
- Included a mock API client so the entire pipeline can be built, run, and tested end to end without spending any API credits.
Tech stack: Python, Gemini API (tool-calling), Streamlit, Flask, n8n, Lovable, JSON Schema
Bone Fracture Detection and Diagnosis System
An AI diagnostic tool that finds and highlights fractures in X-rays
- Combines three different model types to both classify and precisely localize fractures, deployed as a usable tool for real-time uploads.
What it does
- Given an X-ray image, the system classifies whether a fracture is present, segments the exact fracture region, and overlays a bounding box for quick visual confirmation, then serves the result through a live upload interface aimed at medical professionals.
What I did
- Combined three complementary model types in one pipeline: a ResNet classifier (92% accuracy), a U-Net segmentation model for precise fracture region outlining, and YOLOv11 for detection, which alone boosted detection accuracy by a further 10%.
- Curated and labeled the medical imaging dataset myself using Labelbox and Roboflow, ensuring clean, high-quality training data rather than relying on a pre-cleaned public set.
- Used OpenCV to render accurate visual fracture overlays on top of the original X-ray for easy interpretation.
- Estimated the system cuts diagnostic review time by roughly 20% compared to manual review.
- Built and deployed a Streamlit-based interface supporting real-time X-ray uploads, so the tool is usable by someone with no ML background.
Tech stack: Python, PyTorch, ResNet, U-Net, YOLOv11, OpenCV, Labelbox, Roboflow, Streamlit, Google Colab (GPU)
Drug-to-Drug Interaction Prediction Model
Predicting risky drug combinations using graph and sequence models
- Pairs Graph Neural Networks with Recurrent Neural Networks to flag harmful drug interactions before they happen.
What it does
- The model represents drugs and their properties as a graph structure and combines this with sequential modeling to predict whether a given pair of drugs is likely to produce a harmful interaction, aimed at supporting safer prescribing decisions.
What I did
- Combined Graph Neural Networks (GNNs), to capture relational structure between drug properties, with Recurrent Neural Networks (RNNs), to capture sequential/sequence-based patterns, rather than using either architecture alone.
- Improved prediction accuracy by 12% over baseline approaches.
- Built and evaluated the model in PyTorch, using scikit-learn for supporting analysis of drug properties.
- Framed the project around a concrete pharmaceutical research use case: flagging harmful combinations early, rather than treating it as a purely academic exercise.
Tech stack: Python, PyTorch, Graph Neural Networks (GNNs), Recurrent Neural Networks (RNNs), scikit-learn
Two-Stage Deep Learning Pipeline for Aerial Building Footprint Detection
Automatically detecting building footprints in aerial imagery, end to end
- A two-stage computer vision pipeline built from scratch in PyTorch: a classifier decides whether a building is present, then a segmenter outlines exactly where it is.
What it does
- Given an aerial photo, the system first uses an EfficientNet-B0 classifier to determine whether the tile contains a building, then passes positive tiles to a DeepLabV3+ segmentation model (with a MobileNetV2 encoder) that produces a pixel-level mask of the building's footprint. The two-stage design avoids wasting the segmenter's effort on empty tiles, keeping the pipeline efficient.
What I did
- Designed and trained both stages of the pipeline from scratch, rather than using an off-the-shelf end-to-end segmentation model, to keep inference cheap on tiles with no building present.
- Reached 96.8% classification accuracy and 0.645 IoU / 0.785 Dice on the segmentation stage, training end-to-end on free Google Colab GPUs.
- Carried out error analysis to understand exactly where and why the model fails (e.g. dense urban clutter, ambiguous rooftops), rather than only reporting aggregate accuracy.
- Built a Streamlit demo so the pipeline could be tried live on new images, not just evaluated offline.
Tech stack: Python, PyTorch, EfficientNet-B0, DeepLabV3+, MobileNetV2, Streamlit, Google Colab (GPU)
Frequently asked questions
Have questions? Find more information here.
Adithya is based in Osnabrück, Germany.
Adithya speaks the following languages: Marathi (Native), English (Advanced), Hindi (Advanced), German (Elementary).
Adithya has at least 0 years of experience. During this time, Adithya has worked and for 1 company. The average length of individual experience is 0 months. Note that Adithya may not have shared all experience and actually has more experience.
In recent years, Adithya has worked for Vehicle Classification and Detection using Neural Networks.
Adithya is most experienced in industries like Automotive and Information Technology.
Adithya is most experienced in business areas like Information Technology and Research and Development.
Adithya holds a Master in Cognitive Science from Universität Osnabrück and a Bachelor in Computer Science & Engineering from St Joseph Engineering College (VTU).
The availability of Adithya needs to be confirmed.
Daily rate distribution
The rates shown represent the typical market range for freelancers in this position based on recent contracts on our platform.
Average rates for similar positions
Rates are based on recent contracts and do not include FRATCH margin.
The average daily rate is the mean of all daily rates from recent contracts of comparable freelancers on our platform.
The median daily rate is the middle value of all daily rates — half of comparable freelancers charge less, half charge more. Unlike the average, it is barely affected by outliers.
Calculated based on our freelancers’ daily rates as of 18 Nov 2026. Actual rates may vary depending on seniority level, experience, skill specialization, project complexity, and engagement length.
Similar freelancers
Discover other experts with similar qualifications and experience
Kiran Kumar K.
Last position:
Applied NLP: Word-Level Encoding for Smarter Event Predictions at University of Siegen
- Engineered 12+ Seq2Seq models (LSTMs/GRUs) to train an AI model, supporting AI Agent Evaluation Analyst and online projects in complex systems.
- Conducted 15+ experiments to improve forecasting accuracy by 28%, applying analytical thinking and testing models for QA and edge case coverage.
- Researched 20+ papers as a consultant to guide Large Language Model design, ensuring logical implications and domain of expertise alignment.
- Saved 40% compute time via model compression with reusable PyTorch framework, aiding developers and writers to suggest refinements and improve policies.
Ateet B.
Last position:
AI Engineer at MASX AI
Strategic transition into AI Engineering through intensive mentoring and project execution.
Developed MASX AI, an agentic AI platform integrating LangGraph, AutoGen, and RAG for geopolitical forecasting and real-time ETL.
Designed and delivered functional AI prototypes for prospective clients showcasing applied expertise in multi-agent systems, real-time data pipelines, and LLM integrations.
Malik O.
Last position:
AI Scientist at SQPS
- Intelligent Document Processing Engine with OCR & Computer Vision: Architected the end-to-end AI pipeline at the core of the SQPS GenAI document-processing product, which turns scanned paper and PDF batches into organised, searchable documents, combining image pre-processing, OCR and layout analysis for mixed-quality scans. Technologies: Python, OpenCV, Tesseract / PaddleOCR, PyTorch, FastAPI
- Automatic Document Splitting & Classification: Developed models that detect where one document ends and the next begins inside a single multi-document scan, and classify each one by type (invoices, contracts, letters, reports) to enable fully automated filing. Technologies: Python, PyTorch, scikit-learn, Hugging Face Transformers
- Information Extraction & Intelligent File Naming with LLMs: Implemented an LLM-based GenAI extraction system that reads dates, names and reference numbers from unstructured documents and uses them to generate consistent, rule-based file names, with configurable naming conventions per organisation. Technologies: Python, LangChain, OpenAI, SentenceTransformers, SQL
Abhishek N.
Last position:
Fullstack Developer at DAMALO GmbH
- Own full-stack development of an AI-native enterprise platform built on TypeScript, React, Vite, tRPC, Hono, and PostgreSQL, delivering AI-powered consulting workflows to B2B clients.
- Designed and shipped a multi-agent AI system using ReAct framework and Claude skills-style workflow patterns, including an intelligent PM assistant with rich system prompts, slash commands, tool integrations, and streaming chat UI.
- Architected an LLM evaluation framework: rubric-based LLM-as-judge, golden datasets, regression testing, and automated quality gating — ensuring consistent AI output quality at scale.
- Integrated LangFuse for end-to-end LLM tracing, conversation replays, and evaluation pipelines, enabling data-driven prompt optimisation that reduced token costs and response variance.
- Built with Drizzle ORM, pgvector, and knowledge graphs for structured data access, semantic search, and relationship-aware AI reasoning across the platform.
- Led TanStack React Query migration across the application — replacing manual state management with centralised caching and automatic refetching, reducing data-fetching boilerplate significantly.
- Practiced AI-native development throughout: Claude Code, Codex, Perplexity SDK, and LLM-assisted testing across the full development lifecycle. Deployed on Vercel + Azure ACA with Biome for linting/formatting.
Mathew D.
Last position:
Data Science Expert and AI Strategist at Freelancer
- Built an API to ingest, clean, translate, and index EU tenders documents in Neo4j, enabling hybrid search with RAG and Cypher queries via a Streamlit dashboard
- Deployed the API on AWS Lightsail container services with CI/CD automation via GitHub Actions, ensuring stability through pytest unit and integration tests
- Designed and developed a comprehensive online course on data analysis using ChatGPT for professionals and learners, creating instructional videos and interactive Jupyter notebooks
- Utilized OBS and professional audio equipment to ensure high-quality video and audio content
- Led a CRM data normalization and cleaning project visualized via a Sankey diagram to aid customer understanding and pipeline development
- Implemented and validated a genAI-driven web crawling strategy on AWS, ensuring data quality, scalability, and CRM data augmentation
Tobias J.
Last position:
Design of an AI-Agent-Based ERP System
- Design of an LLM-based agent system to control the ERP software
- Development of agent workflows with LangGraph and PydanticAI
- Planning interfaces between business logic and language models
- Planning agent orchestration
- Prototype development and demonstration
Tools: Python, Pydantic, React, LangChain, LangGraph, Linux
Experts recently working on similar projects
Freelancers with hands-on experience in comparable project as a
Nearby freelancers
Professionals working in or nearby Osnabrück, Germany
