Johnson Controls AI Agents Cut RFP Processing to One Minute

Johnson Controls' agentic AI system for RFP processing is an enterprise tool that automates the extraction of specifications from request for proposal documents to support commercial workflows.

Executive Summary

Johnson Controls has introduced an AI agent to manage the processing of request for proposal documents in its commercial operations. The company operates in the building technologies sector where responding to RFPs is a regular part of business development. The AI agent is capable of extracting relevant sections from documents up to 300 pages long in approximately one minute. This represents a significant reduction from the previous timeline of several days for manual review. The primary outcome is the return of 8 to 10 hours of seller time per week, enabling the sales team to dedicate more effort to client engagement and proposal customization rather than document parsing. This time recovery allows sellers to pursue more opportunities or to spend additional time on relationship building with potential clients. The efficiency gain can lead to improved response quality as more time is available for tailoring proposals to specific customer needs. In a competitive market, such advantages can translate to higher win rates and increased revenue over time. The deployment is part of a strategic effort to leverage AI for workflow optimization in enterprise settings.

In addition to the time savings for sellers, Johnson Controls has undertaken a staffing transformation in its digital organization. Over an 18-month period, the proportion of internal staff increased to 70 percent from a previous level of approximately 30 percent. This insourcing effort complements the AI deployment by building internal capabilities to develop and maintain such systems. The AI agent operates in a parallel manner, allowing it to handle multiple RFPs for up to ten sellers at the same time. These changes provide a concrete example of how AI can be applied to specific enterprise processes to yield measurable improvements in efficiency and resource allocation. The shift supports greater alignment between digital initiatives and core business objectives while reducing reliance on external providers for ongoing development work.

The quantified win for Johnson Controls lies in the combination of time recovery for the sales force and the shift toward greater internal control over digital functions. Executives can see direct benefits in terms of increased seller capacity without proportional increases in headcount. The initiative is part of a larger rebuild of commercial, service, and customer support workflows. While the current AI enhancements have delivered notable results, the company indicates that substantial additional gains remain possible. This positions the firm to continue optimizing its operations through further AI integration. The approach demonstrates how targeted AI applications can address operational bottlenecks in document intensive industries.

Key metrics comparing operations before and after AI agent deployment at Johnson Controls.

Metric	Before	After
RFP Processing Time	Days	One minute
Weekly Seller Time on Value-Added Tasks	10 hours target baseline	8-10 hours improvement achieved
Digital Staffing Composition	30 percent internal	70 percent internal
RFP Handling Mode	Serial	Parallel for up to 10 sellers

1. The first step involved identifying the RFP processing bottleneck where manual review consumed excessive seller time in the commercial workflow.
2. The second step focused on developing the agentic AI capable of extracting sections from 300 page documents in one minute.
3. The third step enabled parallel processing to support up to ten sellers simultaneously without delays.
4. The fourth step included measuring the resulting time savings of 8 to 10 hours per week per seller.
5. The fifth step encompassed the organizational shift to 70 percent internal digital staffing over 18 months to support ongoing AI development.

Background and Context

The building technologies sector involves complex projects that often require detailed bidding processes through RFPs. Companies must review extensive documentation to ensure compliance with specifications and to develop competitive quotes. This process has historically been time consuming and labor intensive, involving teams of sellers and technical experts who spend hours or days sifting through pages of requirements. Johnson Controls, as a major player in this space, faces these challenges regularly as part of its commercial activities. The drudgery associated with RFP handling can limit the number of opportunities a seller can pursue and delay response times, potentially impacting win rates and revenue generation. In such environments, any reduction in administrative overhead directly supports the ability to scale operations without equivalent growth in personnel costs.

Prior to the AI deployment, the manual nature of RFP processing meant that sellers spent a significant portion of their time on administrative tasks rather than on strategic selling activities. This allocation of time represented an opportunity cost for the organization. The digital organization at Johnson Controls was also heavily reliant on external resources, with roughly 70 percent of the team being external prior to the insourcing initiative. This structure could introduce challenges in knowledge retention and alignment with internal priorities. The decision to invest in agentic AI for this workflow reflects a strategic response to these operational realities. Addressing both the technology and the talent composition allows for a more cohesive approach to digital transformation.

Enterprise AI applications in commercial workflows have gained attention as organizations seek to automate repetitive tasks. In the case of Johnson Controls, the focus on RFP processing targets a bottleneck that affects the front end of the sales cycle. The podcast discussion with Vijay Sankaran on Enterprise AI Innovators provides insight into the motivation behind the project. The goal was to increase the amount of time sellers could spend on high value activities, with an initial target of moving from ten hours to 20 hours per week on such tasks. The AI enhancements have contributed to progress toward that objective. This context highlights the practical business drivers behind the adoption of agentic systems in established enterprises.

The context also includes the broader effort to rebuild workflows across commercial, service, and customer support areas. By addressing the RFP stage, Johnson Controls aims to create efficiencies that propagate through the customer engagement process. The staffing shift to 70 percent internal supports this by ensuring that the teams responsible for AI development and maintenance are aligned with the company's long term objectives. This combination of technology and organizational change provides a model for other enterprises facing similar document heavy processes. The sector wide implications include potential standardization of AI assisted bidding practices that could reshape competitive dynamics over time.

What's New in Detail

The new development at Johnson Controls is the deployment of an AI agent specifically designed for the RFP-to-quote pipeline. This agent automates the initial step of identifying and pulling out relevant sections from incoming RFP documents. The capability to complete this extraction in one minute allows for rapid turnaround and enables sellers to begin their analysis sooner. The parallel processing feature means that the system can handle multiple documents concurrently, supporting a team of up to ten sellers without queuing delays. This represents an advancement over traditional serial processing methods that would handle one request at a time. The innovation lies in applying agentic AI to a high volume, repetitive task that previously required substantial human effort.

The AI agent is already in use and delivering results in the form of time savings. Sellers are reporting the recovery of 8 to 10 hours per week that can now be redirected to other activities. This is part of a phased approach where the current implementation is described as less than 25 percent of the way to the full vision. The system integrates into the existing commercial workflow, providing outputs that sellers can use to prepare quotes more efficiently. The focus on agentic AI allows for autonomous operation on the defined task of document extraction. Early adoption within the sales organization has validated the approach and encouraged further exploration of similar applications.

Details from the Enterprise AI Innovators coverage indicate that the agent was built to address a specific pain point identified by the sales team. The feedback from users highlights the dramatic reduction in processing time from days to one minute. This new capability changes the dynamics of how RFPs are managed, allowing for greater volume handling without additional staff. The insourcing of the digital team has likely facilitated the rapid development and iteration on this AI solution by bringing expertise in house. The combination of these elements marks a notable update to the company's commercial processes.

Technical Specifics

From a technical perspective, the AI agent utilizes agentic capabilities to navigate and parse large PDF or document files containing RFP content. It identifies sections that are relevant to the quoting process based on predefined criteria or learned patterns from previous RFPs. The one minute processing time for a 300 page document demonstrates the efficiency of the underlying models and algorithms employed. The parallel operation is achieved through scalable infrastructure that allows multiple instances of the agent to run simultaneously for different sellers. This technical design ensures that the system can accommodate varying workloads without degradation in performance.

The system is designed to run in the background and deliver extracted information directly to the relevant seller or team. This integration reduces the need for manual intervention in the initial review phase. The ability to process multiple RFPs at once addresses the common scenario where a sales team receives several opportunities in a short period. According to the source material from Enterprise AI Innovators, the agent was developed internally following the shift to a majority internal digital organization. This internal development likely allows for customization to the specific needs of Johnson Controls' commercial processes. The architecture supports ongoing refinement as more data from real world usage becomes available.

While specific model architectures or training data details are not disclosed in the available reports, the performance metrics indicate a robust implementation. The agent operates autonomously on the extraction task, freeing human resources for higher level decision making. The parallel nature ensures that the system scales with the number of active sellers, up to the limit of ten concurrent processes mentioned. This technical approach aligns with the goal of augmenting rather than replacing the sales workforce. The emphasis on speed and concurrency addresses key requirements in fast paced bidding environments.

Market and Stakeholder Implications

For executives in similar sectors, the Johnson Controls example illustrates the potential for AI to impact sales productivity directly. By reducing the time spent on RFP processing, companies can potentially increase the number of bids submitted or improve the quality of responses through more time for customization. The time savings of 8 to 10 hours per week per seller translate to a significant annual productivity boost when aggregated across a sales team. This can contribute to revenue growth if the additional time leads to more closed deals or better client relationships. The approach also highlights the importance of measuring outcomes in concrete terms such as hours returned rather than abstract efficiency gains.

The staffing shift to 70 percent internal has implications for cost structure and capability building. External staffing often carries higher costs and less control over priorities. By bringing the digital team in house, Johnson Controls gains the ability to iterate on AI solutions more quickly and align them closely with business needs. This model may be of interest to other enterprises looking to balance AI adoption with organizational development. The combination of AI tools and internal talent creates a sustainable approach to digital transformation that can be adapted across industries.

Stakeholders including sales leaders and IT executives can draw lessons from the quantified outcomes. The reduction in processing time from days to one minute highlights the speed advantage of AI agents in document analysis tasks. The parallel processing capability ensures that the benefits scale with team size. For the building technologies sector, where RFPs are a standard part of business, this type of automation can provide a competitive edge in responsiveness. The reported progress toward the goal of 20 hours of seller time on value added activities suggests ongoing opportunities for further optimization. Executives should consider how similar bottlenecks exist in their own operations and whether agentic AI offers a viable solution.

The implications extend to workforce augmentation, where AI handles the drudgery and humans focus on judgment and creativity. This can improve employee satisfaction by reducing repetitive work. For the C-suite, the case demonstrates a clear path from AI deployment to measurable business outcomes without requiring massive headcount increases. The 18 month timeline for the staffing change shows that such transformations are achievable within a reasonable period. Other companies may consider similar dual focus on technology and talent to achieve comparable results. The overall effect is a more agile commercial operation capable of handling increased volume with existing resources.

Expert Reactions

Vijay Sankaran, Chief Digital and Information Officer at Johnson Controls, has provided commentary on the AI initiative in discussions with Enterprise AI Innovators. His remarks emphasize the practical impact on seller time and the potential for further development. The executive notes the specific improvement in RFP handling as a step toward broader workflow enhancements. This reaction underscores the value placed on the AI agent by leadership responsible for digital strategy. The perspective from the CDIO offers a grounded assessment of both achievements and remaining opportunities.

this would take us days to do. And you’ve built an agent that can do this for us in a minute. And by the way, it’s not serial, it’s parallel. I can run ten different reps at once.Vijay Sankaran, Chief Digital and Information Officer, Johnson Controls

Another statement from Sankaran highlights the partial nature of the current achievements. He indicates that the team is not even 25 percent of the way to the full potential of the AI enhancements. This perspective suggests that while the initial results are positive, the organization views the current deployment as a foundation for additional capabilities. The reaction from the CDIO provides a realistic view of the progress and the roadmap ahead for Johnson Controls' AI efforts in commercial workflows. Such commentary is valuable for peer executives evaluating similar investments.

What's Next

Looking ahead, Johnson Controls plans to build on the current AI agent to achieve greater time savings for sellers. The stated goal includes moving toward 20 hours of seller time per week on value added activities, with the current AI contributing 8 to 10 hours of the improvement. Since the implementation is described as less than 25 percent complete, additional features and refinements are expected in the coming periods. The parallel processing and extraction capabilities can be expanded to cover more aspects of the RFP and quoting process. Continued iteration will likely focus on increasing the scope of automation while maintaining the accuracy required for commercial use.

The insourcing of the digital organization to 70 percent internal provides a strong base for continued development of AI solutions. Internal teams can focus on identifying new use cases within the commercial, service, and customer support areas. The success with the RFP agent may inspire similar applications in other document intensive workflows. Executives at peer companies can monitor the evolution of this system for insights into scalable enterprise AI deployments. The internal capability also supports faster response to emerging needs as the business environment evolves.

Overall, the trajectory suggests sustained investment in agentic AI to further optimize operations. The combination of technology deployment and organizational change positions Johnson Controls to realize ongoing benefits in productivity and efficiency. As more of the potential is unlocked, the company may see compounding effects on its commercial performance. This forward looking approach serves as a reference for other organizations considering AI in their core business processes. The case underscores the value of starting with targeted applications that deliver clear returns before expanding the scope.