Sainyam Kapoor

CEO

Seattle, Washington, United States9 yrs 2 mos experience

Most Likely To SwitchHighly Stable

Key Highlights

Senior AI leader at Amazon's AGI Foundation.
Led development of Nova models for real-world intelligence.
Improved answer rates by over 3% in AI systems.

Stackforce AI infers this person is a leader in AI and machine learning, specializing in model evaluation and development.

Contact

Skills

Core Skills

Model EvaluationMachine LearningNeural NetworksNatural Language ProcessingTrading SignalsData Analysis

Other Skills

BenchmarkingParaphrasingMedium Frequency TradingReport GenerationOutlier DetectionPythonResearchProgrammingNetwork SecurityData StructuresStrategic PlanningCC++LinuxRed Hat Linux

About

I am a senior AI leader building frontier-scale AGI models. At Amazon’s AGI Foundation, where I work on Nova models, I lead end-to-end model evaluation and benchmarking across real-world use cases and academic benchmarks - measuring model capability as systems progress from random weights to real-world intelligence.

Experience

9 yrs 2 mos

Total Experience

2 yrs 3 mos

Average Tenure

6 yrs 7 mos

Current Experience

Amazon

3 roles

Senior Manager

Promoted

Oct 2025 – Present · 8 mos

Leading a team of MLEs evaluating Nova models and developing alternative inference stacks to enable faster experimentation.

Model EvaluationBenchmarkingMachine Learning

Software Development Manager

Jul 2022 – Oct 2025 · 3 yrs 3 mos

Senior Software Development Engineer

Nov 2019 – Jul 2022 · 2 yrs 8 mos

Facebook

AI Resident

Oct 2018 – Oct 2019 · 1 yr · Greater New York City Area

Part of the First Residency class at Facebook AI Research(FAIR)

Amazon

Software Engineer

Sep 2017 – Sep 2018 · 1 yr · Cambridge, United Kingdom

I worked with Alexa's knowledge federation team.
What's that?
Federation team was responsible for ranking possible answers and paraphrasing the query to something we can answer.
How did you do that?
We developed NN models to order the answers based on the context and annotations from Internal workflow.
Any project you are proud off?
I developed a paraphrasing system which improved the overall answer rate by more than 3% and increased customer satisfaction by 86 basis points.

Neural NetworksParaphrasingNatural Language Processing

Trexquant investment lp

Global Alpha Researcher

Jan 2017 – Aug 2017 · 7 mos

Building trading signals for medium frequency trading.

Trading SignalsMedium Frequency Trading

Goldman sachs

Summer Analyst

May 2015 – Jul 2015 · 2 mos · Bangalore

Ranking Outlier VMs based on severity, which is defined by violations.
Automated Report Generation for affected VMs(neighboring tenants) and Applications
Designed Generic Outlier tool to identify and report Outliers based on business rules
Analyzed and studied metrics such as Swap-In and Swap-out of the hypervisor, to detect business rules for outliers and demonstrated a correlation between metrics and incident reports.

Data AnalysisReport GenerationOutlier Detection