MAYANK SONY | 21CH3FP28

Indian Institute of Technology, Kharagpur

CHEMICAL / FINANCIAL ENGG.
sonymayank11@gmail.com | 9123695425
EDUCATION
Year Degree/Exam Institute CGPA/Marks
2026 B.Tech+M.Tech Dual Degree 5Y IIT Kharagpur 8.38 / 10
2021 ISC Delhi Public School Newtown, Kolkata 92.75%
2019 ICSE The Assembly of God Church School, Kolkata 96%
INTERNSHIPS
Live Project Intern | Futures First Private Limited Oct'23 – Dec’23
Building Dynamic Pairs Trading Strategy using Kalman Filter
• Implemented the distance approach using sum of squared deviation and co-integration method running ADF test for finding out pairs
• Introduced Kalman Filters for dynamic tracking of the hedge ratio between equities of pairs to keep the spread stationary
• Achieved a CAGR of 9.59% and Sharpe Ratio of 1.41 by taking positions when the half-life of the Z-score breaches entry and exit Z-score
AI/ML Intern | AlgoBulls Technologies Private Limited Jun'23 – Nov’23
Custom Chatbot powered by Generative AI models for inhouse trading software
• Developed GPT 3.5-based chat model by utilizing Langchain and ChromDB for streamlining vector storage and retrieval optimization
• Achieved an improvement of 30% in code generation accuracy by analyzing multiple generative AI models like Llama 2 and Deepseek
• Improved the code efficiency by a significant 15% by formulating a better searching algorithm using the FAISS and BM25 retriever
Strategy Intern | FluxGen Sustainable Technologies Pvt. Ltd Jan'23 – Mar’23
Conducted extensive Market Research for FluxGen (a B2B startup in Water Sustainability)
• Comprehensive research to identify FluxGen’s top 250 clients in India based on ESG Score, Industrial Belt, and Profitability
• Rated each company based on metric defined on bases of ESG rating, Profitability, Water-Stressed Regions, and Sector
• Prepared an extensive client database along with the probability factor for becoming a potential client, for 300+ companies
Research Intern | Professor Dmitri Kaplun | Saint Petersburg Electrotechnical University LETI May'23 – Jul’23
Construct a 3D mesh model directly from CT images
• Conducting research to identify suitable datasets, resulting in selection of Spleen and ShapeNet dataset for Mesh model development
• Analysed and reviewed various Mesh generation algorithms, ultimately implementing Mesh R-CNN, to directly utilize the 3D images
PROJECTS
Indian Sign Language Recognition System | OpenCV, TensorFlow, Python, Kotlin May'22 - Jun'22
Course project | Professor Dipankar Debnath and Titash Mondal | IIT Kharagpur
• Built an ISL Recognition System with Voice Assistance Integration by leveraging transfer learning on pretrained TensorFlow Lite models
• Enhanced systems performance by augmenting a dataset of 2000 images covering 20 distinct ISL signs, ensuring robustness and accuracy
• Achieved outstanding real-time ISL recognition accuracy rate of 88%, facilitating seamless communication through precise interpretation
COMPETITIONS
Gold | Inter-IIT Tech Meet 12.0 | DevRev’s AI Agent 007| IIT Kharagpur Nov'23 - Dec'23
• Proposed a method called "Reimagining Tooling as Coding" (RTaC) that treats tool usage by large language models as a coding task
• Utilized training techniques, including LoRA and PEFT, to ensure robust performance and efficient deployment with minimal surprises
• Conducted experiments with three pipelines for open-source models: few-shot prompting, tool memorization, and proposed RTaC
• Implemented RTaC on models like CodeLlama-7B and DeepSeek 1.3B, achieved close to GPT-4 level performance with fewer parameters
Bronze | General Championship Ad Design 2023 | IIT Kharagpur Feb'23 - Mar'23
• Conducted extensive Market Research and devised market a entry strategy for fitness company and revamped their pricing strategy
• Compared virtual fitness companies using DCF valuation and strategic frameworks based on its financials, advantages, and strategies
POSITIONS OF RESPONSIBILITY
Executive Head | Quant Club, IIT Kharagpur Jun'22 - Present
• Part of a team of 40+ quant enthusiast promoting Quantitative Finance among 5k+ students at IIT Kharagpur through diverse initiatives
• Organized Summer of Quant, featured coursework on Quantitative Finance and Algorithmic Trading had a footfall of 3.5k+ participants
• Working towards creating alphas by developing and backtesting strategies using technical analysis with statistical and ML algorithms
Vice President | Senior Member | Junior Member | ShARE, IIT Kharagpur Jun'22 – Present
• Heading the local chapter of global leadership program, held extensive knowledge sessions for 30+ students on Marketing & Finance
• Generated revenue with YOY growth of 200% through start-up projects by landing three new clients in Cleantech & Fintech verticals
• Overhauled existing structure, improved output by 50% through a sector-based team division, boosted media outreach to over 70%
COURSEWORK INFORMATION
IIT KGP: Programming and Data Structures (Theory+Lab) | Linear & Matrix Algebra | Advanced Calculus | Transform Calculus | Economics
MOOCS: STATS 101 | Data Structures & Algorithms | Machine & Deep Learning Specialization (Deeplearning.AI) | LangChain for LLM
Application Development (Deeplearning.AI) | Zerodha Varsity | Bloomberg Market Concepts | Options 101 by Akuna Capital
SKILLS AND EXPERTISE
C | C++ | Python | Java |Matplotlib | pandas | NumPy | Langchain | SQL | seaborn | scikit-learn | SciPy | statsmodel | OpenCV | Pytorch
EXTRA CURRICULAR ACTIVITIES
• Represented Meghnad Saha Hall of Residence, IIT Kharagpur in field of technology as Secretary to ensure smooth execution of events