Aditya Mahakali
Solving Problems with AI
Summary
I'm a full-stack AI Engineer, building impactful solutions by harnessing the power of Machine Learning and Generative AI to deliver real value to users and clients.
Innovative and analytical problem-solver with keen interest in artificial intelligence and machine learning. Well-versed in Python programming and machine learning. Committed to leveraging AI technologies to drive impactful solutions and optimize processes.
Skills
Artiificial Intelligence Engineer
Advance
Machine Learning, Deep Learning, Generative AI, Natural Language Processing, Computer Vision, AI Search, Multimodal Models, Agents, MCP
Backend Development
Intermediate
Spring Boot, Django, flask, nodejs, Fast API
Database Mangement Systems
Intermediate
SQl, NOSQl, GraphQl, Elastic, Milvus, weaviate, Chroma, Neo4j
Devops
Intermediate
Linux, git, docker, openshift and ansible
Profiles
github
linkedin
leetcode
Medium
Experience
IBM
Bangalore
AI Engineer
August 2023 – Present
•
Piloted full-stack AI solutions using Watsonx for 12+ clients across healthcare, banking, and telecommunications, resulting in 95% client satisfaction based on post-MVP feedback surveys.
•
My expertise spans the full AI engineering stack—from architecting robust backend systems and developing secure APIs to building intuitive and user-friendly frontends.
•
Worked on many internal productivity tools for developer as well as managerial roles like OnePagerPro, Vulnerability Scanner, etc.
•
Participated in research work over RAG Techniques, Knowledge Graphs, Multi Agentic Collaboration.
Hughes Systique Corporation
January 2023 – August 2023
Software Development Intern
•
Built PoCs using LLMs on private business data by integrating Hugging Face transformers with vector stores (Milvus, Chroma, Weaviate) and LangChain-inspired tools like AI chat interfaces and web retrievers.
•
Completed full-stack training and developed a Spring Boot + React + Docker project, published on GitHub.
Examinations Cleared
GATE CS
2023
IIT
UGC Net CS Qualified
2023
UGC
IIT JAM Mathematics
2021
IIT
JNUEE
2021
JNU
Cleared JNUEE exam in 2021
Education
Jawaharlal Nehru Univeristy
MCA • 7.65/9
Computer Science and Applications
2021 – 2023
Studied Advance computer science concepts with ML and NLP
Central University of Rajasthan
BSC Computer Science • 7.74/10
Computer Science
2018 – 2021
Studied computer science fundamental, statistics as well as grad level mathematics
Publications
SEARCHD
23 October 2024
IEEE
SEARCHD - Advanced Retrieval with Text Generation using Large Language Models and Cross Encoding Re-ranking
Study on How different RAG techniques can deliver greater faithfulness, context relevance and answer relevancy.
Projects I worked on
IBM- Networking Client
Implemented a Custom RAG solution that can answer questions via semantic similarity(vectors) as well as SQL like query. Using advance full-text search and filtering capabilities of Elastic-SQL we were able to deliver great customer value. Also did:
•
Fine-tuning embedding models
•
Developed high-accuracy query classifier
•
Assisted in fine-tuning reranker
IBM- Virtual Assistant Banking MVP
•
Architected and developed a custom end-to-end RAG-based chatbot using Watsonx.ai LLMs and Watson Assistant, embedded in a React UI with features like deep links, card carousels, and map-based location queries.
•
Deployed frontend and backend services as Docker containers on IBM Cloud.
IBM - Product Search and virtual assistant MVP
•
Leveraged generative AI to create summarised product descriptions from FDA documents, significantly improving search relevance for a medicinal drug store.
•
Developed and deployed a RAG-based Virtual Assistant on IBM Cloud, enabling customers to query product information and place orders, enhancing customer support and streamlining the purchasing process.
IBM - Retail NL2SQL
Created a Natural Language to SQL pipeline using watsonx.ai LLMs to answer natural language queries of workers managing orders and sales.
•
Automated Metadata Dictionary Creator
•
Created Custom Disambiguation
•
Integrated with watsonx assistant via Backend APIs in python(FastAPI)
OTHERS Pilots- IBM
•
Marketing Campaign / Image Generation: Using Flux, created a automatic image-prompt improving agent using state space search. check_demo
•
Document Generation : Created similar document based on a given template and a new Idea.
•
Watsonx Challenge Project – VSCode Extension for Code Vulnerability Detection
•
Knowledge graph rag using LLMs over neo4j, we automated ontology creation, triplet extraction and natural language to cypher query.
Personal - Daily Research Paper Summarizer
Leveraging Gemini Created a summarizer agents which will create a summary of all the popular research that happen on that day.
PERSONAL - Med Bot
leverage hypothetical document embeddings to create good quality RAG answers on medical Question Answers
Personal - Guesthouse Management System
I Built this project using Django Backend to create software that enables the users to manage guesthouse at my university.
Personal - Bug Pilot
A simple yet powerful bug-tracking app built with ReactJs for the front end and Springboot for the backend APIs. It provides full CRUD functionality to help you efficiently track and manage bugs in your software development process.
Interests
Chess
Teaching
Certifications
Intro To Python, Algorithmic ToolBox
Coursera
Intro To ML, Intro to deep learning, Intro to Computer Vision
kaggle