Deepesh Mahawar

Associate Partner

Mumbai, Maharashtra, India8 yrs 10 mos experience

Highly Stable

Key Highlights

6 years of experience in Quantitative Finance and automation.
Dual degree from IIT Bombay in Electrical Engineering and Microelectronics.
Proficient in advanced programming languages and HPC.

Stackforce AI infers this person is a Quantitative Finance professional with strong expertise in High Performance Computing.

Contact

Skills

Core Skills

Quantitative Analysis (finance)High Performance Computing (hpc)Machine Learning

Other Skills

CUDAMPIC++Gaussian EliminationNewton-RaphsonLU factorizationSpacetime Supercomputer Cray XC50MNA matrixFFTMotor Current Signature AnalysisLossless Data CompressionMPLAB SIMCPPCPU profilingVLSI

About

A goal oriented individual with demonstrated history of working in Quantitative Finance and automation for 6 years. Strong research professional with degree focused in Electrical Engineeeing (B.TECH) + Microelectronics(M.TECH) Dual Degree from IIT Bombay. Proficient in C++, Python, R, CUDA, MPI, Excel.Contact me at deepeshkumarmahawar@gmail.com, mobile 7738847950

Experience

8 yrs 10 mos

Total Experience

2 yrs 3 mos

Average Tenure

2 yrs 1 mo

Current Experience

Ubs

2 roles

Associate Director

Promoted

Jan 2025 – Present · 1 yr 3 mos · Mumbai, Maharashtra, India

Quant Analyst(Authorized Officer)

Mar 2024 – Feb 2025 · 11 mos · Mumbai, Maharashtra, India

Credit suisse

Quant Analyst(Exempt-Non Officer)

Mar 2020 – Feb 2024 · 3 yrs 11 mos · Mumbai, Maharashtra, India · On-site

Ee iit bombay

Research Scholar

Jan 2018 – Feb 2020 · 2 yrs 1 mo · Mumbai Area, India

D-latch circuit simulation is done using Modified Nodal Analysis and dynamic expressions obtained are evaluated over time using Newton-Raphson iterations.
In each iteration, the linear equations obtained are solved using Gaussian Elimination method.
After setting basic algorithm created variations in the circuit parameters of 10%(which occurs naturally during production) to study the effect on circuit behavior.
Simulated circuit on k40 GPU using Cuda parallelization and Cudastream to vary parameters parallely, performing a total of 40,00,000 simulations, thus obtained a speedup of 149.5 compared to CPU.
To further optimise used LU factorisation in symbolic form using SymPy to generate L and U factors of the MNA matrix in terms of node voltages.
Used Spacetime Supercomputer Cray XC50 system's Pascal GPU for further testing and got additional speedup of 6.7.
Implemented Cuda and MPI combined for further scaling the application on multiple nodes.

CUDAMPIC++Gaussian EliminationNewton-RaphsonLU factorization+4

Iit bombay

Graduate Teaching Assistant

Jul 2016 – Apr 2017 · 9 mos · MUMBAI

Introduction To MEMS (Micro Electro- Mechanical System)
Silicon Pressure sensors, Micromachining, MEMS Fabrication :Integrated Circuit Processes, Bulk Micromachining ,Physical Microsensors, Microactuators: Electromagnetic and Thermal microactuation.
Microelectronics Simulation Lab
Familiarization with TCAD software, CMOS structure design, Compact models and transistor-level circuit simulation.

Forbes marshall

Summer Internship | Analyst | everSENSE Remote Monitoring App

May 2015 – Jun 2015 · 1 mo · Pune Area, India

Tested and Integrated FFT algorithm with Eversense mobile app to monitor rotor and stator conditions in real time
Analysed signatures originated from different faulty motor parts by using Motor Current Signature Analysis
Implemented algorithm for Lossless Data Compression on VIBTRANS (dspPIC33F128G306) using MPLAB SIM
Impact: Improved Eversense user window data refresh time from 6 to minimum 1 sec

FFTMotor Current Signature AnalysisLossless Data CompressionMPLAB SIMMachine LearningQuantitative Analysis (Finance)

Creative concepts

Winter Internship | Asst. Embedded System Engineer - 'Automated Tea Maker'

Dec 2014 – Dec 2014 · 0 mo · Vadodara Area, India

Modeled a Tea Maker using stepper motor, solenoid valve, 16x2 LCD, Hex keypad, hotplate, relay on AT89S52 Ultra Development Kit using 8051 Assembly language.
System can enroll individual preferences for quantity of each ingredients using their respective user IDs.