Vishwas Lathi

Software Engineer

Kanpur Nagar, Uttar Pradesh, India5 yrs 11 mos experience

AI ML PractitionerAI Enabled

Key Highlights

Expert in C++ with a focus on high frequency trading.
Proficient in Machine Learning and neural network models.
Hands-on experience with real-time data processing tools.

Stackforce AI infers this person is a Fintech-focused Software Engineer with strong capabilities in Machine Learning and high frequency trading.

Contact

lathivishwas@gmail.com LinkedIn

Skills

Core Skills

C++High Frequency TradingMachine LearningNeural Style TransferDocument Vectorization

Other Skills

AlgorithmsArtificial Intelligence (AI)BashCComputer ScienceComputer VisionDBSCANData StructuresDatabasesDeep LearningGDBGitGo (Programming Language)HaskellJSON parsing

Experience

5 yrs 11 mos

Total Experience

5 yrs 11 mos

Average Tenure

5 yrs 11 mos

Current Experience

Tower research capital

2 roles

Software Engineer

Jul 2020 – Present · 5 yrs 11 mos · Gurgaon, Haryana, India

Software Engineer Intern

May 2019 – Jul 2019 · 2 mos · Gurgaon, Haryana

● Implemented data fetching and analyzing tools in C++ for maintaining low latency data connections to the exchanges for high frequency trading.
● Implemented efficient JSON parsing techniques to reduce the tick to trade latency of the existing trading platforms by the order of milliseconds.
● Implemented tools for processing and analyzing depth, level and trade messages to maintain a
real-time order book for various markets.

C++JSON parsingdata fetchinghigh frequency trading

New york office, iit kanpur

Machine Learning Intern

May 2018 – Jul 2018 · 2 mos · Kanpur, Uttar Pradesh

● Created an online document vectorization model based on the Distributed Memory paragraph vectors model.
●Created an online clustering model using a fully-online modification of DBSCANoptimised for discrete data.
● Created and trained a Word2Vec model to compare scraped articles using Word Mover’s Distance.
● Implemented a fast neural style transfer model based on Perceptual Losses for Real-Time Style
model.

Machine LearningWord2VecDBSCANneural style transfer