Atin Modi

Senior Software Engineer

Bengaluru, Karnataka, India6 yrs 2 mos experience

Highly StableAI Enabled

Key Highlights

Expertise in Generative AI and scalable systems.
Patented automated testing pipeline for ML.
Strong background in algorithm design and optimization.

Stackforce AI infers this person is a Software Engineer with expertise in AI and algorithm development across HealthTech and Bioinformatics.

Contact

Skills

Core Skills

Machine LearningDeep LearningAlgorithmsData StructuresSoftware DevelopmentWeb Development

Other Skills

RNNSVRLinear RegressionHyper-Parameter tuningDynamic ProgrammingGenomic Multiple Strings AlignmentGeneralised Hidden Markov ModelSAGEObject-oriented programmingAlgorithm DesignAnalytical TheoryHTMLCSSJavaScriptMySQL

About

Software Engineer in the Azure OpenAI team, working on Generative AI platforms and Azure OpenAI services, with a focus on scalable systems, reliability, and responsible AI in production. Previously at Oracle Unity / Marketing Cloud, I shipped GenAI + ML platform features (including secure “bring your own model” execution on Kubernetes) and built a patented drift-based automated testing pipeline that reduced ML production bugs. I also have research experience collaborating on problems from number theory to statistical inference. Outside work, I’m active in financial markets (F&O, FX/commodities) and build algorithmic strategies using math + ML.

Experience

6 yrs 2 mos

Total Experience

2 yrs 3 mos

Average Tenure

1 yr 6 mos

Current Experience

Microsoft

2 roles

Senior Software Engineer

Mar 2026 – Present · 3 mos · Bengaluru, Karnataka, India · On-site

Software Engineer II

Nov 2024 – Feb 2026 · 1 yr 3 mos · Bengaluru, Karnataka, India · On-site

Oracle

2 roles

Senior Software Engineer

Promoted

Sep 2023 – Nov 2024 · 1 yr 2 mos

Application Engineer

Sep 2020 – Sep 2023 · 3 yrs

Samsung electronics

Summer Intern

May 2019 – Jul 2019 · 2 mos · Noida, Uttar Pradesh, India

Built a recommendation system using RNN (Recurrent Neural Network) to suggest a new workout routine for a user based on his previous performances
Compared and trained SVR, Linear Regression and RNN for the given data which helped in proper Hyper-Parameter tuning to learn data patterns
Successfully Improved the previous recommendation system and is now updated in the Samsung Health Mobile Application and can suggest new workout routines based on current weather, air quality, BMI and previous workout records

RNNSVRLinear RegressionHyper-Parameter tuningMachine LearningDeep Learning

Universität greifswald

Research Assistant

Dec 2018 – Jan 2019 · 1 mo · Greifswald, Mecklenburg-West Pomerania, Germany

Worked on improving a Dynamic Programming Algorithm for Genomic Multiple Strings Alignment of more than 100 strings with each string comprising of one million characters
Successfully developed a scoring scheme and a cost function to align genes (which are basically long strings made from 4 characters) satisfying some constraints with better matching results
Tuned the parameters for Generalised Hidden Markov Model to improve the efficiency of Genomic Multiple Strings Alignment

Dynamic ProgrammingGenomic Multiple Strings AlignmentGeneralised Hidden Markov ModelAlgorithmsData Structures

University of luxembourg

Research Assistant

May 2018 – Jul 2018 · 2 mos · Luxembourg

Developed a user surface in SAGE, an open-source multiple-software package, for the mathematical program Bianchi.gp making use of the object-oriented structure available in SAGE
Added documentation enabling the user to prompt the function descriptions directly from the SAGE terminal
This newly developed program will be included in the next version of SAGE, which will help in solving the problem of finding Bianchi Orbifolds of Small Discriminant

SAGEObject-oriented programmingSoftware Development

The university of hong kong

Research Assistant

May 2017 – Jul 2017 · 2 mos · Hong Kong

Designed an optimized algorithm which computes the ternary sum with a complexity order of O(n*logn) despite the brute force algorithm having a complexity of O(n^3)
Found the conditions on m for which the ternary sum of m-gonal numbers is almost universal using the above algorithm
Found the local obstructions of the problem using both algebraic and analytical theory by representing an integer as a Lattice Coset

Algorithm DesignAnalytical TheoryAlgorithms

Cognizance, iit roorkee

Member of Promotion and Development Team

Dec 2015 – Feb 2016 · 2 mos · IIT roorkee

Helped developed the website of Cognizance 2016 using web technologies like HTML5, CSS, Javascript, Jquery, MySQL, PHP (http://cognizance.org.in/)

Web DevelopmentHTML5CSSJavaScriptjQueryMySQL+1

Creative mammoth

Web Developer

Sep 2015 – Mar 2016 · 6 mos · IIT roorkee

Developed a fully functional website (freelance) for the company using web technologies like HTML, CSS, Javascript, MySQL, PHP (http://creativemammoth.com/).

Web DevelopmentHTMLCSSJavaScriptMySQLPHP