We start with the result you need, cut unnecessary steps, and build AI agents and automation to deliver it. We prepare your team to use and maintain what we build.
You’ll work directly with the founders, Linda and Allan.
5 people, 1 weekBefore
1 person, 1 dayWith StrataEdge
Each case describes what we built and links to the full write-up.
In production. We continue to improve it with BCG.
BCG needed an AI agent that returns structured results people can review. StrataEdge and two engineers built a working version in four weeks. It has review checkpoints where a person confirms each result before it is used.
Read the BCG caseProof of concept, completed in 14 weeks
A service member’s treatment record draws on separate operational, treatment, and purchased-care systems. We built a multi-cloud data ecosystem that brings those records into governed data products with full lineage. AI agents analyze the records and answer questions across documents, and service members review and certify their record in a portal.
Read the Defense Health Agency caseOne million records. The proof of concept used synthetic health records.
Implemented by DigiKey
We developed the organization design and development plans and delivered leadership development and generative AI training. DigiKey implemented the new roles, jobs, and development plans.
Read the DigiKey caseTechnology teams in the United States and India
Our software compares bordereaux, commissions, premiums, and claims with the contracts behind them. It shows the evidence for each check and sends discrepancies to an auditor for review. We built it for finance, audit, and delegated authority teams at carriers, MGAs, and program administrators.
At an insurance carrier that works with more than 250 MGAs, quarterly auditing fell from five people for a week to one person for a day. Manual checks found that 99% of outputs met audit requirements, compared with 90% before.
You can hire us for one service or several.
AI agents and automation
We remove unnecessary steps, build AI agents, and automate the work that remains. We show supporting evidence and route exceptions to people for review.
We built a working AI agent for BCG in four weeks and took it to production in three months.
AI agents and automationData and technology modernization
We connect fragmented records, preserve their sources, and improve the software and data flows your teams rely on.
For the Defense Health Agency, we built a multi-cloud data ecosystem and AI agents that assemble treatment records from separate health systems.
Data and technology modernizationAI strategy and governance
We assess the work, data, and full cost of delivery and operation. You get investment priorities, measures of success, and clear responsibility for decisions.
Linda managed GEICO’s $1.2B technology budget as CTO. Allan’s company FamilyLink reached more than 100 million users. Both work directly with clients.
AI strategy and governanceOrganization design and training
We define roles, career paths, and leadership responsibilities, and deliver leadership development and AI training. Engage us for this work alone or as part of an AI or technology project.
DigiKey implemented new roles, jobs, and development plans for its technology teams in the United States and India.
Organization design and trainingMost clients start with a review or a proof of concept. We agree on scope, cost, and the result we will measure before we start. Linda and Allan work directly with your team throughout.
2–4 weeks
We review the work, data, and full cost with your leaders. You get a ranked list of projects with their costs, including work that needs no AI and work you should stop.
4–8 weeks
We build a working version against your real examples and measure it against a result we agree on first. For BCG, this took four weeks.
Scoped after the proof of concept
We connect the application to your systems, set up monitoring and human review, and train the people who will run it. BCG’s agent was in production three months after we started.
Scoped per team
We define roles, career paths, and leadership responsibilities, and train leaders and staff to use AI. DigiKey implemented the roles and plans we developed.
Most of our work is in insurance finance and audit, mortgage operations, and government health records. We have also worked with BCG in professional services and DigiKey in electronics distribution.
For carriers, MGAs, and program administrators
Our software compares bordereaux, commissions, premiums, and claims with contracts. As GEICO’s CTO, Linda rebuilt its insurance data ecosystem.
Insurance finance and auditFor agencies and prime contractors
For the Defense Health Agency, we built a data ecosystem and AI agents that assemble service members’ treatment records from separate health systems.
Healthcare and government recordsFor lenders, brokers, and credit unions
Our team built LoanFlow, which cut the time to clear a loan for closing from 30 days to under 24 hours and lowered back-office costs by 75%.
Mortgage operationsWe completed a proof of concept for the Defense Health Agency under contract HT003826PE001. We also join larger programs as a teaming partner to a prime contractor. Linda Apsley is the agency contact.
Conversations and articles from Linda and Allan about planning, building, and running AI.
Allan Carroll with Zhen Lu, CEO and co-founder of Runpod. They discuss where data should live, replaceable model choices, evaluation, and human approval when agents act quickly. Video, 32 minutes.
Watch the conversationYou’ll work directly with them. They staff each project from the thousands of engineers they have worked with.
Book a 30-minute callOr email info@strataedge.ai. We reply within one business day.
Managing Partner & CEO · Senior Advisor at BCG
Linda leads AI strategy, technology investment, and organizational change. At Citi, she was Managing Director and Head Architect for finance and risk technology. At Capital One, she led data engineering and moved financial and risk data to the cloud.
Managing Partner & CTO
Allan leads AI and product engineering. He has built products and engineering teams for more than 20 years, and his companies raised over $100M. His AI work began with graduate research at the University of Washington in 2006.