Summer 2026

Research Intern

AI Frontiers, Reasoning & Agentic Models

Posted on 4/3/2026

Microsoft

Microsoft

10,001+ employees

Develops software, OS, and cloud services

Compensation Overview

$32.37 - $63.52/hr

Company Historically Provides H1B Sponsorship

Redmond, WA, USA

In Person

Must be physically located at a Microsoft worksite in Redmond, WA for the duration.

Bachelor's, Master's, PhD

Category
AI & Machine Learning (3)
, ,
Required Skills
LLM
Neural Networks
Machine Learning

Get referred to Microsoft

See people who can refer or advise you

Requirements
  • Accepted or currently enrolled in a bachelor's, master's or PhD program in Computer Science or related STEM field.
  • At least 1 year of experience with deep learning and large language model training.
  • Research Interns are expected to be physically located in their manager’s Microsoft worksite location for the duration of their internship.
  • Submit a minimum of two reference letters for this position as well as a cover letter and any relevant work or research samples. After you submit your application, a request for letters may be sent to your list of references on your behalf.
Responsibilities
  • During the 12-week internship, Research Interns are paired with mentors and expected to collaborate with other Research Interns and researchers, present findings, and contribute to the vibrant life of the community.
  • Research internships are available in all areas of research, and are offered year-round, though they typically begin in the summer.
Desired Qualifications
  • Accepted or currently enrolled in graduate program in Computer Science or related STEM field.
  • Proven expertise in language model pre-training, post-training, or reinforcement learning.
  • Proven publication record in top-tier conferences.
  • Demonstrated ability to develop original research and perform hands-on research in a collaborative and dynamic environment.

Microsoft develops software, devices, and cloud services. Windows is an operating system that runs on personal computers, Office provides productivity apps, and Azure offers cloud computing and developer tools. The company differentiates itself with a large, integrated ecosystem of software, devices, and services, plus long-standing partnerships with PC makers and a broad enterprise footprint. Its goal is to put a computer on every desk and in every home, and to extend that reach through cloud services, professional networking (LinkedIn), and gaming.

Company Size

10,001+

Company Stage

IPO

Headquarters

Redmond, Washington

Founded

1975

Get referred to Microsoft

See people who can refer or advise you

Simplify Jobs

Simplify's Take

What believers are saying

  • Azure grew 43% in fiscal Q4 2026, beating estimates and accelerating AI consumption.
  • Microsoft forecast about 45% Azure growth in fiscal Q1 2027, sustaining momentum into 2027.
  • Copilot unification rolled out in August 2026, expanding paid AI adoption across consumer and enterprise users.

What critics are saying

  • Reuters said the FTC escalated scrutiny in February 2026 over Microsoft’s licensing and bundling.
  • Microsoft cut 4,800 jobs in July 2026, including 3,200 in Xbox, showing churn.
  • UK regulators reopened cloud licensing probes on March 31, 2026; breakup pressure is real.

What makes Microsoft unique

  • Azure crossed $100 billion annual revenue in fiscal 2026, validating Microsoft’s scale.
  • Microsoft bundles Windows, 365 Copilot, Teams, and Azure into one enterprise AI stack.
  • Microsoft’s partner ecosystem, from Darwinbox to Sentinel integrations, deepens switching costs across workflows.

Help us improve and share your feedback! Did you find this helpful?

Benefits

Health Insurance

Dental Insurance

Vision Insurance

401(k) Company Match

Professional Development Budget

Conference Attendance Budget

Flexible Work Hours

Remote Work Options

Growth & Insights and Company News

Headcount

6 month growth

0%

1 year growth

0%

2 year growth

0%
Yahoo Finance
Sep 3rd, 2026
Microsoft reveals Azure generated $29.4B last quarter and $101.9B in fiscal year

Microsoft disclosed Azure's revenue for the first time, revealing the cloud platform generated $29.4 billion in the most recent quarter and $101.9 billion for the fiscal year ending 30 June. Azure accounted for roughly half of Microsoft Cloud's $59.3 billion quarterly revenue and grew 43% year-on-year. The company is restructuring its reporting segments into two divisions: Agents and Infra, plus Devices and Consumer. Microsoft Cloud's commercial remaining performance obligations reached $678 billion. The disclosure ends years of opacity around Azure's financial performance, giving investors clearer metrics to assess whether artificial intelligence investments are capturing market share. Microsoft shares traded at $512.72, approximately 12% below analyst estimates of fair value at $581.42.

Yahoo Finance
Sep 2nd, 2026
Microsoft shifts cloud story from migrations to AI workloads as Azure posts 43% growth

Microsoft has shifted its cloud narrative from customer migrations to AI-driven workloads, signalling a strategic pivot in how the company generates growth. Once dominated by stories of enterprises moving operations to Azure, management now emphasises AI platform adoption and compute consumption. Azure posted 43% growth, significantly outpacing the broader Intelligent Cloud segment's 32% increase. The migration story, previously central to earnings calls, has gone quiet — not due to failure, but maturity. The company now highlights Azure OpenAI Service usage and hyperscale capacity expansion. This shift suggests Microsoft has successfully transitioned from selling one-time infrastructure moves to recurring, consumption-based AI revenue. The Intelligent Cloud segment generates roughly $137.8 billion annually, growing at 30%. The narrative change concentrates Microsoft's growth prospects and risks squarely on AI adoption rates.

Associated Press
Sep 2nd, 2026
Teradata integrates enterprise AI with Microsoft OneLake via Apache Iceberg

Teradata has integrated its Autonomous Knowledge Platform with Microsoft OneLake, allowing enterprise customers to run Teradata's AI directly on data stored in OneLake without moving or duplicating it. Built on Apache Iceberg standards, the integration enables users to query OneLake tables in place using standard Iceberg APIs, with Microsoft Entra ID handling cross-platform authentication. Enterprises can now join and analyse data across both environments without consolidating it into a single system or rebuilding governance frameworks. The integration eliminates the need for separate ETL pipelines to synchronise data copies across systems, reducing latency, governance gaps, and storage costs. Data engineers spend less time maintaining pipelines, whilst AI teams can begin feature engineering without waiting for data replication. Read access to Microsoft OneLake tables via Teradata is now available. The integration will be demonstrated at the European Microsoft Fabric + SQL Community Conference in Barcelona from 28 September to 1 October 2026.

Associated Press
Sep 2nd, 2026
Darwinbox expands Microsoft partnership to apply Frontier Tuning to HCM platform

Darwinbox has expanded its partnership with Microsoft to integrate Microsoft Frontier Tuning into its human capital management platform. The collaboration makes Darwinbox among the first enterprise software companies to apply this technology, which individualises workforce interactions and improves over time. The integration works through Microsoft Copilot and Microsoft Teams, following the launch of Darwinbox Cortex. The partnership aims to deliver an HCM system that becomes more valuable with use, rather than depreciating after deployment. Founded in 2015, Darwinbox serves over 1,400 organisations and 4.5 million employees worldwide. The company has been recognised as a Challenger in Gartner's Magic Quadrant for HCM platforms. Investors include KKR, Partners Group, OTPP, Microsoft, and Salesforce Ventures.

Innovation News Network
Sep 2nd, 2026
Microsoft invests $5M in ALERTCalifornia AI system to detect wildfires earlier

Microsoft is investing $5 million in ALERTCalifornia, a UC San Diego programme using AI and over 1,300 high-definition cameras to detect California wildfires early. The funding includes $2 million for technology development and a $3 million Azure grant. The system analyses more than seven million images daily, identifying smoke and potential fires before emergency services receive reports. During its first season in 2023, it detected over 1,200 fires, spotting incidents before 911 calls more than 30% of the time. Some wildfires were detected as much as 2.5 hours before the first emergency call. The AI distinguishes actual fire activity from environmental changes like clouds or fog. Emergency personnel verify detections and use multiple camera angles to assess developing incidents, improving response times and decision-making for California's increasingly challenging wildfire conditions.

INACTIVE