iTromsø · IBM, VC2
10 minutes per day — time spent reviewing potential stories
Independently reportedMedia & entertainment · Document processing · NLP · In production
Document processing
IBM · Rohirrim AI
Vendor-reportedTechnology & softwareUnited StatesScaledEnterpriseLLM
Vendor-reported. The customer is named and the numbers are quoted from the source page, but the account comes from the vendor. No independent confirmation.
Reported by rohirrim.ai (vendor self-report). Checked against the source page on 2026-09-05. 2 of 2 figures below appear on that page word for word.
The problem
IBM faced difficulties managing their vast proposal content library and streamlining their RFP response process, leading to ineffective knowledge base searching, SMEs struggling with simplifying technical concepts, lengthy new team member training, and limited proposal handling capacity.
What was deployed
- Vendor
- Rohirrim AI
- Products
- RohanRFP
- Technique
- LLM
- Build or buy
- Bought off the shelf
- Deployment
- SaaS
- Data used
- IBM’s historical content, proposals (both winning submissions and those that scored well technically), past performance documentation, and artifacts (including resumes, white papers, and RFI responses).
- Scale
- across IBM Public, Americas, and Global divisions
What changed
Each row is quoted from the source. Figures we could not find on the page in those words are marked — they are kept, not deleted, so you can judge them.
60 minutesproposal research and drafting timefrom 6 days
“Reduced proposal research and drafting time by 90%, from 6 days to 60 minutes”
Quoted word for word from the source
under a weeknew employee training periodfrom 6 months
“Decreased new employee training period from 6 months to under a week”
Quoted word for word from the source
multiple proposals concurrentlyproposal handling capacity for solution architects
“Enabled solution architects to handle multiple proposals concurrently”
Quoted word for word from the source
consistent high-quality outputquality and consistency of output
“Achieved consistent high-quality output across all business units”
Quoted word for word from the source
“What previously took six days of research and discovery can now be accomplished in about 60 minutes of carefully crafted prompting,”
Difficulties and limits
- Traditional content management systems couldn’t answer nuanced questions like “How do we handle DevSecOps for the Department of Defense in 2023?”
- Subject matter experts (SMEs) struggled to translate their technical expertise into well-written proposals
- New team members required months of training to become familiar with existing content
- Solution architects were limited in how many proposals they could handle simultaneously
- Maintaining consistent quality across different business units proved difficult
- general-purpose AI tools couldn’t meet their specialized needs
- We couldn’t risk our intellectual property being exposed to public AI models.
- While the platform doesn’t directly create slides
Sources
- rohirrim.ai — vendor self-report, 2025-07-03
Similar deployments
Amplitude · Wolfia
4X — throughput of security questionnaires
Vendor-reportedTechnology & software · Document processing · AI agents · Scaled
Adyen · Rossum
70 — automation rate
Vendor-reportedBanking & finance · Document processing · NLP · Scaled · 2023
Skaylink · Moterra
dropped by around 80% — NIS2 preparation time
Vendor-reportedTechnology & software · Document processing · LLM · Scaled
OM Pharma · DrugCard
3 minutes — review time per article
Vendor-reportedPharma & biotech · Document processing · NLP · In production
AMPREP (Amalgamations Repco) · Cimplify.AI
100% — invoice automation
Vendor-reportedManufacturing · Document processing · NLP · In production