Senior AI Engineer
I'm a Senior AI Engineer with 4+ years of experience and Founder of DevFuzzion.com
Senior AI Engineer
Aditya Sakpal
Brief initial presentation of myself
Available for work
Hello I am Aditya Sakpal
Senior AI Engineer and Founder of DevFuzzion Based in Mumbai, India .
I'm Aditya Sakpal, a dedicated AI specialist based in the vibrant city of Mumbai, India. Open to work with cool startups and even with established business out there .
Pipeline Architect
RAG
AI Agents
Chatbots
LLM Finetuning
Prompt Engineering
NLP Tasks
Minor Projects
Mistral-7B Finetuned
Real-Time Call Intelligence
AI Swarm
Sentimental Analysis
How it works
Process Is Everything
Simple, streamlined process is what get's you results
I am with you in every step
alongside you at each step for seamless experience
Education
Education
"Memorize → Vomit on Paper → Forget → Repeat”
Certifications
Certifications
Showcase of some of my certifications as a proof of my excellence
Why choose me
Why me as AI Engineer
Why select me as your AI Engineer
Experience
Clients Record
4+ years of experience
6
companies
60 Lakhs +
Revenue generated so far for other businesses via AI
4.8
Average Rating
JuteQ Inc , Canada
Senior AI Engineer
July.2024-Sep.2025
Created AI Agents using Langgraph and PHIData, integrated frontend and backend with Livekit, and enabled custom agent creation through the Eliza framework. Enhanced prompts, improved query routing, and leveraged Langsmith for debugging along with advanced Prompt Engineering. Designed multiple agentic workflows delivering business outcomes, including retrieving JSON schema outputs from OpenAI and streaming its API with FastAPI integration. Built an estimate tool for resolving roofing issues, designed a complete chatbot pipeline for a DAO (Decentralized Autonomous Organization), and implemented Redis caching for memory optimization. Scraped data from platforms like Telegram and Discord, extracted insights from BRSR reports and financial statements using LLMs, and managed Supabase DB schemas with custom RPC functions. Built APIs with FastAPI, integrated authentication with Clerk and Swix, contributed to frontend development using Next.js, and deployed applications on Azure.
Infigos Solutions , US
AI Developer
July.2024-Sep.2025
Created an AI system for predicting anomalies in finance data to identify fraud, with the model developed in Red Hat Decision Manager. Developed backend and frontend for IFRS17 and IFRS4 banking applications, utilizing Docker for running Red Hat Decision Manager.
QNext AI , US
AI Developer
Nov.2024-Jan.2025
Extracted info from brsr reports and financial statements using LLMs for finance application |Managed Supabase DB schemas and created RPC functions | Built APIs with FastAPI, integrated Clerk and Swix authentication,and deployed on Azure|Contributed in frontend(Next js)
Hooman Digital LLP , India
AI Developer
July.2024-Sep.2025
Designed entire pipeline for the chatbot system for DAO(Decentralized Autonomous Organization). Created AI agents using PHIData and made them interact with social media via Eliza framework Enabled custom agent creation | Scrapped data from various data sources including telegram and discord channels | Built 80% pipeline |Implemented redis caching for memorization | Streamed Open AI’s API
Scorelab & C2SI (For Google Summer for CODE) , SriLanka
Open Source Contributor
Oct. 2023 – April. 2024
Resolved frontend issues to improve UX with React.js, engineered APIs using Firebase Functions, and optimized data handling with Firestore and Firebase Storage, while utilizing the Firebase Emulator for seamless local testing. Designed system architecture for Codelabz, developed ER diagrams, and documented contributions.
BeyondChats , India
AI Developer
Jan.2023-July.2024
Implemented a chat system for building emotional connect with IVF patients by creating APIs in Flask. Developed auto novel generation and SEO metadata generation pipelines using Azure OpenAI for LLM calls. Improved the RAG pipeline and built a web crawler/scraper leveraging Selenium, BeautifulSoup, Scrapy, and Playwright, while also testing multiple use cases on MS Copilot Studio. Automated data insertion into Pinecone and fine-tuned a BERT model for NER tasks on resumes, significantly improving entity recognition accuracy. Crafted a chatbot using the RASA framework and React.js, leading to reduced response times. Additionally, streamlined data analysis workflows through web scraping with Selenium, data preprocessing, and backend integration with FastAPI.
IGSRS Solutions, India
AI Developer
Oct.2022-Dec.2022
Fine-tuned BERT model for NER task on resumes, significantly improving entity recognition accuracy. • Crafted a chatbot using the RASA framework and ReactJS, resulting in a noticeable decrease in latency. Streamlined data analysis by implementing web scraping using Selenium & integrating FastAPI .
HFT Solutions , India
Full Stack AI Developer
July. 2022 – Oct. 2022
Implemented AWS Rekognition for improved accuracy and reliability, with logging for efficient monitoring. Built a missing persons web application for Rajasthan Police and a Telegram bot using FastAPI, with the UI developed in React.js. Established database schemas in Postgres, configured logging, and used WinSCP for SSH.
Finplex Solutions, India
ML Engineer
April 2022 – July 2022
Developed a chatbot for IPPB, utilizing QLoRA for fine-tuning, significantly improving customer satisfaction. Integrated the OpenAI API with Flask and tested multiple LLMs for better accuracy, while creating the UI using React.js.
FAQ Section
Questions, Answers
Get quick answers to your most pressing questions
Joe Rogan
PowerfulJRE
I recently interviewed Mr. Aditya and surprise, surprise, he actually seems to be an intelligent guy (wasn’t on my bingo card) here are the questions I threw at him.
What’s your experience with scaling AI systems for real-world usage?
Scaling involves optimizing both inference and infrastructure. I’ve implemented Redis caching for reducing redundant API calls, containerized services with Docker, and deployed apps on Azure. I also use Supabase and Postgres for structured data, ensuring scalability for both storage and compute.
How do you keep up with the rapidly evolving AI ecosystem?
How do you approach improving the accuracy of LLM responses in a production environment?
If AI could replace one daily task for you, what would you happily hand over?
You’ve worked on chatbots and agents if you had to design a sarcastic chatbot version of yourself, what’s the first feature you’d add?
You’ve mentioned creating AI agents using Langgraph and PHIData. Can you explain how you structured these agents and what business outcomes they achieved?
You fine-tuned a BERT model for NER on resumes. What improvements did you achieve, and how did you measure success?
You built an estimate tool for roofing issues how did AI help in this scenario?
You’ve integrated Clerk and Swix authentication in FastAPI and deployed on Azure. What challenges did you face with authentication and deployment?
