Adept
Adept automates end-to-end enterprise workflows using AI agents that interact with software applications.
Adept is an enterprise AI platform that automates complex, multi-step workflows across software applications by directly perceiving and acting on user interfaces like a human would. It uses proprietary models to locate elements on webpages, reason about documents, and execute end-to-end business processes—from data extraction and form population to license application processing. The platform is designed for enterprises seeking to offload repetitive tasks while keeping humans in the loop for critical decisions.
Problem solved
Enterprise teams waste significant time on manual, repetitive, end-to-end workflows across multiple software applications that could be automated.
Target customer
Enterprise organizations with high-volume, repetitive, multi-step workflows across multiple software systems; particularly those in operations, finance, and compliance teams.
Founders
D
David Luan
CEO
Co-founder of Google Brain, VP of Engineering at OpenAI (2017-2020) where he led research on language models, supercomputing, and reinforcement learning; shipped GPT, CLIP, and DALL-E; BS in Applied Mathematics and Political Science from Yale.
N
Niki Parmar
Co-founder (departed November 2022)
Co-author of the Transformer paper (Attention Is All You Need); left to pursue another startup.
A
Ashish Vaswani
Chief Scientist (departed November 2022)
Co-author of the Transformer paper (Attention Is All You Need); left to pursue another startup.
Z
Zach Brock
CEO (current)
Head of Engineering, promoted to CEO; 15+ years building software products and leading teams; successful entrepreneur and former CEO.
Funding history
Series A
$65M
April 2022
Led by Greylock, Root Ventures
· Scott Belsky (Behance), Howie Liu (Airtable), Dara Khosrowshahi (Uber), Chris Re (Stanford), Lip-Bu Tan (Cadence), Andrej Karpathy (Tesla), Nathan Benaich (Air Street), David Ha, Jaan Tallinn, Sarah Meyohas, Kevin Kwok
Series B
$350M
March 2023
Led by General Catalyst
· Spark Capital, Addition, Greylock, Atlassian Ventures, Microsoft, Nvidia, Workday Ventures, Caterina Fake, Frontiers Capital, PSP Growth, SV Angel, A.Capital
Total raised:
$415M
Industries
Pricing
Not publicly available; product is in private beta. Contact required for pricing information.
Integrations
Web applications and software systems via direct screen interaction; no specific third-party integrations disclosed.
Tech stack
jQuery (JavaScript libraries)
core-js (JavaScript libraries)
Google Analytics (Analytics)
Google Font API (Font scripts)
Nginx (Web servers)
OpenResty (Web servers)
Varnish (Caching)
Google Workspace (Email)
Google Hosted Libraries (CDN)
Google Tag Manager (Tag managers)
Webflow (Page builders)
Lever (Recruitment & staffing)
Website
Competitors
UiPath
Broader RPA platform focused on process automation; Adept specializes in AI-driven agentic workflows with direct screen perception.
OpenAI
General-purpose large language models; Adept builds narrower, task-specific models designed as software interfaces to mitigate bias issues.
DeepMind/Inflection AI
Similar research direction in AI agents for computer interaction; Adept focused on production enterprise automation.
Anthropic
General AI research; Adept specializes in enterprise workflow automation through software interaction.
Why this matters: Adept represents a new category of enterprise automation powered by AI agents that interact with software like humans do, rather than traditional RPA scripting. The founding team's pedigree (Google Brain, OpenAI, Transformer authors) and substantial funding ($415M) signal serious conviction in agentic AI for enterprise automation, though recent founder departures to Amazon raise questions about the company's long-term vision.
Best for: Enterprise operations, finance, and compliance teams automating high-volume, multi-step workflows across multiple software systems without extensive custom development.
Use cases
Cross-Site Shipping Verification
Adept checks shipping availability across hundreds of vendor websites simultaneously to compile accurate delivery plans instantly. This eliminates manual checking of each site and accelerates fulfillment planning for logistics operations.
Contract Data Extraction and System Update
Automatically extracts key information from PDF contracts and documents, accurately updates internal systems with the extracted data, and sends confirmation emails upon completion. Reduces manual data entry errors and accelerates contract processing from days to minutes.
Compliance-Driven License Application Processing
Processes license applications while adhering to complex business logic rules, keeping humans in the loop for final submission decisions. Ensures regulatory compliance while dramatically reducing the time spent on repetitive form population and validation steps.
Alternatives
UiPath
Mature RPA platform with broader process mining capabilities; choose if you need enterprise-scale automation across legacy systems without AI-driven intelligence.
Automation Anywhere
RPA competitor with strong cloud automation; choose if you prioritize established vendor support over cutting-edge AI agents.
Blue Prism
Enterprise RPA with governance focus; choose if compliance and process control are primary requirements over AI reasoning capabilities.
FAQ
What does Adept do? +
Adept is an enterprise AI platform that automates complex, multi-step workflows by perceiving and acting on software applications like a human would. It uses proprietary models to locate screen elements, reason about documents, and execute end-to-end business processes while keeping humans in the loop for critical decisions.
How much does Adept cost? +
Pricing is not publicly available. Adept is currently in private beta and requires contacting the company directly for pricing information and access.
What are alternatives to Adept? +
UiPath is a mature RPA platform with broader automation capabilities. Automation Anywhere offers cloud-based process automation. Blue Prism specializes in enterprise governance-focused automation. These are more established than Adept but lack its AI-driven reasoning capabilities.
Who uses Adept? +
Target customers are enterprises with high-volume repetitive workflows in operations, finance, and compliance. Specific named customers have not been publicly disclosed, though Adept reports deployments in mission-critical production environments.
How does Adept compare to UiPath? +
Adept uses AI agents that directly perceive screens and reason about tasks, while UiPath is a traditional RPA platform requiring explicit workflow scripting. Adept is more specialized in AI-driven automation but less mature; UiPath has broader process mining and governance tools for large enterprises.
What happened to Adept's founders? +
Co-founders Niki Parmar and Ashish Vaswani (both Transformer paper authors) departed in November 2022 to start another company. CEO David Luan and most co-founders joined Amazon in August 2024; Zach Brock, former Head of Engineering, became CEO of the remaining team (~33% of staff).
Tags
AI agents
workflow automation
RPA
enterprise automation
agentic AI
software interaction
process automation