Summary
Timeline
Skills
Work History
Education
Overview
Work Availability
Hugging Face
Awards
Recent Projects
Hi, I’m

Saksham Jain

Pune
Great knowledge comes from humblest of origins
The Man Who Knew Infinity

Summary

Dynamic IT professional with 6 years of experience in IT consulting, esteemed for delivering over 20 projects across diverse sectors including Finance, Web3, Sports, E-commerce, and Gaming. My expertise encompasses system design, cloud computing, backend development, machine learning, and deep learning, with specialized skills in applying fine-tuning and quantization techniques to enhance AI model efficiency and performance. I have held pivotal leadership roles, such as Technical Head and Technical Architect, inspiring synergistic teamwork and consistently achieving superior results. As COO, I successfully expanded a UK-based company's Indian operations from 4 to 150 employees, showcasing my capabilities in strategic planning, agile methodologies, and executing projects with precision. Renowned for a results-oriented mindset, I am committed to delivering impactful, business-aligned solutions that drive growth and innovation.

Timeline

Data scientist

Freelancer
08.2024 - Current

Technical Head

TECH ALCHEMY PVT LTD
01.2023 - 01.2024

Chief Operating Officer

TECH ALCHEMY PVT LTD
01.2023 - 01.2024

Technical Architect & Team Coordinator

TECH ALCHEMY PVT LTD
11.2021 - 01.2023

Software Developer

TECH ALCHEMY PVT LTD
03.2021 - 11.2021

Software Developer (contractual)

ZoAi Technology
01.2021 - 05.2021

Software Developer

Codium.co
04.2020 - 02.2021

Blockchain Developer

BuyUcoin
01.2019 - 01.2020

Northern India Engineering College

B Tech from EEE
08.2014 - 08.2018

Skills

Technical Designs

undefined

Work History

Freelancer

Data scientist
08.2024 - Current

Job overview

  • Developed and deployed AI-driven solutions for various clients, specializing in LLM fine-tuning, Retrieval-Augmented Generation (RAG), and prompt engineering
  • Designed and optimized machine learning models for diverse applications, including natural language processing (NLP), data analytics, and automation
  • Worked closely with stakeholders to deliver scalable AI solutions, improve model performance, and integrate AI capabilities into business workflows
  • Used MLflow to track experiments results and used DVC pipelines in model deployment
  • Evaluated and compared LLM outputs to identify inconsistencies and provided structured input for Reinforcement Learning with Human Feedback (RLHF) training
  • Trained ML model to identify cheaters, achieving over 90% accuracy in detection
  • Trained ML regression models to find the probability of game being unfair and a separate classification model to classify it
  • Created a Vertex AI batch prediction pipeline using Gemini, along with ML model training and deployment using vertex

TECH ALCHEMY PVT LTD

Technical Head
01.2023 - 01.2024

Job overview

  • Completed over 20 projects across Finance, Web3, E-commerce, Sports, and Hyper Casual games, encompassing full development lifecycle including technology selection, database design, initial setup, and ongoing maintenance
  • Led the data science team for training sentiment analysis of telegram community bots using bagging Random Forest and naive bayes
  • Completed 3 POC of 'CustomGPT' using RAG with Llama-2 7B, GPT 3.5 with pinecone for storing of source documents
  • Closely worked with CTO and implemented better technical practices among organization
  • Includes GitLab to Bitbucket migrations, AWS Cloud best practices
  • Led the data team which completed analysis on Web3 Wallet Transactions, Hyper Casual Games and Casino to increase the retention of user, decrease churn rate and prevent users from cheating (in Games)
  • Implemented advanced AWS cloud strategies to improve security, enhanced fault tolerance, optimized traffic management, increased scalability, cost efficiency, and higher system reliability
  • Led a cross-functional data analysis team for Web3 and Web3 gaming projects, focusing on data-driven insights to guide product development and strategy

TECH ALCHEMY PVT LTD

Chief Operating Officer
01.2023 - 01.2024

Job overview

  • Oversaw day-to-day operations across multiple departments (Technical, Human Resources, Finance, Legal & Operations), ensuring timely delivery of projects and Smooth Daily Operations
  • Led complete India Team which includes Technical Leads, Developers, Quality Assessment Engineer, Human Resources, Project Manager etc
  • Setup Indian Pune office from Scratch to fully functioning
  • Led HR Team and assisted in recruiting, developing, and managing employees to ensure organizational effectiveness and compliance with labor laws

TECH ALCHEMY PVT LTD

Technical Architect & Team Coordinator
11.2021 - 01.2023

Job overview

  • Collaborated with clients to determine project specifications and scope and help team with Software development lifecycle and delivered high-quality solutions within tight deadlines while adhering to budget constraints
  • Proactively monitored system performance trends using Firebase Crashlytics, New Relic and AWS console to anticipate potential issues before issues escalated into larger problems
  • Architected and implemented microservices architecture for scalable and maintainable codebase
  • Leveraged AWS services such as EC2, S3, RDS, SES, Lambda, and Cloud Formation to build and maintain cloud-native applications
  • Utilized Agile methodologies to ensure timely delivery of quality software solutions
  • Designed and implemented RESTful APIs using Nest.js and Node.js to handle complex business logic and data transactions
  • Designed and optimized MongoDB and SQL schemas to support efficient data storage and retrieval
  • Utilized caching strategies (e.g., Redis, in-memory caching) to reduce latency and improve data retrieval times

TECH ALCHEMY PVT LTD

Software Developer
03.2021 - 11.2021

Job overview

  • Developed and maintained scalable back-end services for healthcare applications using Node.js (Express), ensuring high performance and reliability
  • Designed and developed conversational agents using Google Dialog flow to enhance patient engagement and recommend exercises and activities

ZoAi Technology

Software Developer (contractual)
01.2021 - 05.2021

Job overview

  • Developed Python and Tkinter application for system integrated with physical key holder device, leveraging TCP/IP protocol for communication
  • Used Redis (PUB/SUB) for storing history of physical device on database
  • Led pharmacy e-shop project which uses OCR to read prescription of medicine

Codium.co

Software Developer
04.2020 - 02.2021

Job overview

  • Worked with Flutter to develop for money lending agency
  • Created flutter libraries to support Legacy SOUP APIs
  • Developed Inventory management system back end using Python (Django) for Tim Horton's Thailand

BuyUcoin

Blockchain Developer
01.2019 - 01.2020

Job overview

  • Developed and deployed smart contracts for issuing certificates using Solidity and Remix
  • Created analytics and admin APIs for cryptocurrency platform using Python (FLASK)
  • Created Backend using Python (DJANGO) for project which helps deploy smart contract and interact with it with help of APIs

Education

Northern India Engineering College

B Tech from EEE
08.2014 - 08.2018

University Overview

Cleared GATE-2018 examination.

Overview

6
years of professional experience
20

Projects delivered

4
years of post-secondary education
Availability
See my work availability
Not Available
Available
monday
tuesday
wednesday
thursday
friday
saturday
sunday
morning
afternoon
evening
swipe to browse

Hugging Face

Hugging Face

Awards

01/01

Recent Projects

  • Lorahost, A platform hosting multiple fine-tuned LoRA adapters for open-source LLMs with minimal GPU usage (almost same as single base Model), powered by RunPod GPUs, AWS Lambda backend, Cognito user pool, and API Gateway for secure, rate-limited API access. API endpoints mirror OpenAI's for seamless migration.
  • Pokémon Image Classifier, A model designed to identify Pokémon (1025 classes) from images up to the 9th Generation. It employs a Vision Transformer (ViT), paired with a linear classification head utilizing SoftMax activation and optimized with categorical cross-entropy loss.
  • Mongo Query Generator, Created a synthetic dataset of MongoDB queries and prompts based on the top 10 industries that leverage MongoDB for analytical purposes. Identified relevant queries and prompts for each industry and generated synthetic data using ChatGPT and DeepSeek APIs. Used Unsloth for Training and fine tuning purpose.
  • RAG-INSTRUCT, Refined an open-source dataset for a RAG-specialized model using the Gemini API, improving data quality and mitigating hallucinations. Trained LLaMA 3.2 (3B), LLaMA 3.1 (8B), Mistral 7B, and Mistral 8x7B on RunPod GPUs with custom training scripts. Leveraged Unsloth, Transformers, and SFT Trainer for efficient model training and finetuning.
Saksham Jain