> initializing neural core

BRIJESH ADESHARA000%

Ahmedabad, Gujarat, India

BRIJESH
ADESHARA

Applied AI & Full Stack Developer

I build systems that see, reason and act — from real-time vision pipelines to retrieval-grounded agents.

Full Stack

Where it started — and still the foundation everything else ships on.

17 skills

AI / Machine Learning

Retrieval-grounded agents, vision pipelines, and models trained end to end.

14 skills

Data Science

The M.Tech track — experiment tracking, evaluation, and reproducible pipelines.

8 skills

Automation

Wiring the pieces together so they run without me.

7 skills

scroll

01 / 04

Full Stack

Where it started — and still the foundation everything else ships on.

  • React
  • TypeScript
  • JavaScript
  • Tailwind CSS
  • Next.js
  • Node.js
  • Python
  • FastAPI
  • Flask
  • REST APIs
  • PostgreSQL
  • MongoDB
  • Redis
  • Supabase
  • Firebase
  • Docker
  • Git

01Selected work

Six systems, wired together

02Experience & Education

The path here, end to end

  1. AI / ML DeveloperFull-time

    Jun 2026 — Present

    Tassos Consultancy Services Pvt. Ltd.·Ahmedabad, IndiaAI / ML

    • Building an ERP-integrated RAG system — bringing retrieval-grounded AI features directly into the ERP's modules and workflows.
    • Developing an AI-powered recruitment module for the ERP HRMS, automating the hiring pipeline from screening to shortlisting.
    • Embedding LLM-driven capabilities across existing ERP surfaces while keeping answers grounded in the organisation's own data.
  2. AI / ML DeveloperPart-time · Remote

    Nov 2025 — Present

    NGMARS Solutions·RemoteAI / ML

    • Building AI microservices for a research-assistance platform that supports researchers from topic discovery through to publication.
    • Developing Python FastAPI services behind a Node.js / TypeScript API gateway in a containerised microservice architecture.
    • Shipped modules for PDF summarisation and chat, research metadata generation (titles, abstracts, keywords) and emerging-topic trend analysis.
    • Integrated LLM and retrieval pipelines over research papers, with a scalable design aimed at institutional use.
  3. Applied AI & Full Stack DeveloperInternship

    Jan 2025 — Jun 2025

    Tassos Consultancy Services Pvt. Ltd.·Ahmedabad, IndiaAI / ML

    • Developed and integrated LLM-powered chatbot prototypes for business inquiry and customer-support use cases using Streamlit and external AI APIs.
    • Implemented conversational AI for client-facing platforms, including voice-based interactions and agent workflows using tools such as ElevenLabs.
    • Designed and optimised ERP system modules, databases and business logic while contributing to frontend development and responsive UI improvements.
    • Created high-fidelity application UI/UX workflows in Figma — wireframes, responsive layouts and connected screen flows for large-scale applications.
  4. Full Stack Developer (SSRIP)Research Internship

    Jun 2024 — Sep 2024

    Ganpat University·Mehsana, IndiaFull Stack

    • Selected for the Student Summer Research Internship Program (SSRIP).
    • Built HealthAlly, an AI-powered healthcare chatbot for student wellness.
    • Implemented symptom checking, first-aid tracking and appointment booking.
    • Worked in a team of three across Python, Streamlit and MongoDB, from design through to deployment.
  5. Master of Technology, Data Science & Machine Learning

    2025 — 2027

    Rashtriya Raksha UniversityINI·Lavad, Gandhinagar·CGPA 8.00Study

  6. Bachelor of Technology, Information Technology

    2021 — 2025

    Ganpat University·Mehsana·CGPA 8.00Study

03Playground

Don't take my word for it

Two live demos. One asks a retrieval-grounded model about my work; the other runs a neural network entirely in your browser — no server, no API calls.

ask.brijeshRAG · BM25 + embeddings

Ask anything about Brijesh — his projects, stack, experience or availability. Answers are retrieved from his résumé and GitHub, with sources shown.

vision.local0 API calls · on-device

Run a neural network on your own device. Point your camera and it detects objects in real time — the same class of model behind Percepta, with nothing leaving your machine.

~24MB model · loads only when you click

04Certifications

Verified, not just claimed

  • Google / Coursera

    Google Cybersecurity

  • MathWorks

    Deep Learning Techniques in MATLAB for Image Applications

  • HackerRank

    Python

  • HackerRank

    Problem Solving

  • HackerRank

    SQL

  • HackerRank

    Frontend Developer (React)

05Contact

Let's build something

Open to roles and collaborations in applied AI, full-stack and data work. The quickest way to reach me is below — or straight to my inbox.

phone
+91 95106 67912

↳ this form runs through an n8n workflow I built — submissions land in my inbox and trigger an automation.