In the race to bring artificial intelligence into the enterprise, a small but well-funded startup is making a bold claim: the problem holding back AI adoption in complex industries has never been the models themselves.
Contextual AIa two and a half year old company, supported by investors, among others Bezos Expeditions And Bain Capital Venturesunveiled on Monday Agent-composera platform designed to help engineers in aerospace, semiconductor manufacturing, and other technically demanding fields build AI agents that can automate the kind of knowledge-intensive work that has long resisted automation.
The announcement comes at a crucial time for business AI. Four years after ChatGPT sparked a wave of enterprise AI initiatives, many organizations remain stuck in pilot programs and struggling to advance experimental projects to full production. Chief Financial Officers and business unit leaders are becoming increasingly impatient with internal efforts that have cost millions of dollars but produced limited returns.
Dear Keelathe CEO of Contextual AI, believes the industry has focused on the wrong bottleneck. “The model is almost commoditized at this point,” Kiela said in an interview with VentureBeat. “The bottleneck is context: can the AI actually access your own documents, specifications and institutional knowledge? That’s the problem we’re solving.”
Why enterprise AI keeps failing, and what retrieval-enhanced generation should solve
To understand what Contextual AI tries, it helps understand a concept that has become central to modern AI development: retrieval-augmented generation, or RAG.
When large language models like those come out OpenAI, Googlingor Anthropic generate responses, they use knowledge embedded during training. But that knowledge has an end date and cannot include the proprietary documents, technical specifications and institutional knowledge that are the lifeblood of most businesses.
RAG systems attempt to solve this by retrieving relevant documents from a company’s own databases and linking them to the model alongside the user’s query. The model can then base its response on actual business data instead of relying solely on its training.
Kiela helped pioneer this approach during his time as a research scientist at Facebook AI research and later as head of research at Hugging facethe influential open-source AI company. He has a Ph.D. from Cambridge and is an adjunct professor of symbolic systems at Stanford University.
But early RAG systems, Kiela acknowledges, were primitive.
“The early RAG was pretty crude: take an off-the-shelf retriever, hook it up to a generator and hope for the best,” he said. “The pipeline faults worsened. Hallucinations were common because the generator was not trained to remain on the ground.”
When Kiela was founded Contextual AI in June 2023 he wanted to solve these problems systematically. The company developed what it calls a “unified context layer”: a set of tools that sit between a company’s data and its AI models and ensure that the right information reaches the model in the right format at the right time.
The approach has received recognition. According to a Google Cloud case study, Contextual AI has achieved this highest performance in Google’s FACTS benchmark for grounded, hallucination-proof results. The company refined Meta’s open-source Llama models on Google Cloud’s Vertex AI platform, with a specific focus on reducing the tendency of AI systems to fabricate information.
Inside Agent Composer, the platform that promises to turn complex technical workflows into minutes of work
Agent-composer extends Contextual AI’s existing platform with orchestration capabilities: the ability to coordinate multiple AI tools across multiple steps to complete complex workflows.
The platform offers three ways to create AI agents. Users can start with pre-built agents designed for common technical workflows such as root cause analysis or compliance monitoring. They can describe a workflow in natural language and have the system automatically generate a working agent architecture. Or they can build from scratch using a visual drag-and-drop interface that requires no coding.
What sets Agent Composer apart from competing approaches, the company says, is its hybrid architecture. Teams can combine strict, deterministic rules for high-stakes steps – compliance checks, data validation, approval gates – with dynamic reasoning for exploratory analysis.
“For highly critical workflows, users can choose fully deterministic steps to control agent behavior and avoid uncertainty,” Kiela said.
The platform also includes what the company calls “One-click agent optimization”, which takes user feedback and automatically adjusts agent performance. Every step of an agent’s reasoning process can be monitored, and responses are provided with sentence-level quotes that show exactly where the information in the source documents comes from.
From eight hours to twenty minutes: what early customers are saying about the platform’s real-world performance
Contextual AI says early customers have reported significant efficiency gains, although the company acknowledges these figures come from customer self-reporting and not independent verification.
“These come directly from customer reviews, which are approximations of real-world workflows,” says Kiela. “The numbers are self-reported by our customers and describe the before-and-after scenario of Contextual AI adoption.”
The claimed results are nevertheless striking. An advanced manufacturer has reduced root cause analysis from eight hours to 20 minutes by automating sensor data parsing and log correlation. A specialty chemicals company reduced product research from hours to minutes with the help of agents who search patents and regulatory databases. A test equipment manufacturer now generates test code in minutes instead of days.
Keith Schaub, vice president of technology and strategy at Advantagea semiconductor test equipment company, offered an endorsement. “Contextual AI has been an important part of our AI transformation efforts,” said Schaub. “The technology has been rolled out to multiple teams at Advantest and selected end customers, saving significant time on tasks ranging from test code generation to customer engineering workflows.”
The company’s other clients include Qualcommthe semiconductor giant; ShipBoba technology-based logistics provider that claims to have achieved 60 times faster problem resolution; And Nvidiathe chipmaker whose graphics processors power most AI systems.
The eternal enterprise dilemma: should companies build their own AI systems or buy off-the-shelf systems?
Perhaps the biggest challenge Contextual AI faces are not competing products, but the instinct of engineering organizations to build their own solutions.
“The biggest objection is ‘we’re going to build it ourselves,’” Kiela acknowledged. “Some teams are trying. It sounds exciting to do, but it’s exceptionally difficult to do this well at scale. Many of our customers started with DIY and found themselves 12 to 18 months later still debugging retrieval pipelines instead of solving the real problems.”
The alternative – off-the-shelf solutions – comes with its own problems, the company argues. Such tools can be deployed quickly, but often prove to be inflexible and difficult to adapt to specific use cases.
Agent-composer tries to take a middle ground by offering a platform approach that combines off-the-shelf components with extensive customization options. The system supports models from OpenAI, Anthropic and Google, as well as Contextual AI’s proprietary Grounded Language Model, which is specifically trained to stay true to the retrieved content.
Pricing starts at $50 per month for self-service use, with customized enterprise pricing for larger deployments.
“The justification for CFOs is really about increasing productivity and getting them up and running on their AI initiatives faster,” Kiela said. “Every engineering team struggles to hire top engineering talent, so making their existing teams more productive is a big priority in these industries.”
The Road Ahead: Multi-Agent Coordination, Writes, and the Race to Build Composite AI Systems
Looking ahead, Kiela outlined three priorities for the coming year: workflow automation with actual writes on enterprise systems rather than just reading and analyzing; better coordination between multiple specialized agents working together; and faster specialization through automatic learning from production feedback.
“The compound effect is important here,” he said. “Every document you record, every feedback loop you close, the improvements are piling up. Companies that build this infrastructure now will be hard to catch.”
The enterprise AI market remains fiercely competitive, with offerings from major cloud providers, established software vendors, and dozens of startups all chasing the same customers. Whether Contextual AI’s focus on context over models will pay off depends on whether companies come to share Kiela’s view that the fundamental model wars matter less than the infrastructure surrounding them.
But there is a certain irony in the company’s positioning. For years, the AI industry has fixated on building ever bigger, ever more powerful models, pouring billions into the race for artificial general intelligence. Contextual AI makes a calmer argument: for most real-world work, the magic isn’t in the model. It’s about knowing where to look.
#Contextual #launches #Agent #Composer #turn #enterprise #RAG #productionready #agents