Use Case Intelligence RSS Search use cases Search
OpenAI Software / SaaS Asia-Pacific & Oceania

LegalOn Technologies

IT Operations Automation

LegalOn needed to preserve the development-speed gains from Codex while preventing AI spending from growing uncontrollably as model usage expanded across the organization.

Read the original story Explore all use cases
65% reduction in estimated daily costs
Up to approximately 20% cost-efficiency improvement in mature businesses

Reported impact

Business outcome

LegalOn maintained development speed while cutting estimated daily AI costs, improved cost efficiency in mature business units, and created a governance model that ties model choice and future ROI measurement to customer value.

  • Cost Reduction
  • Productivity Improvement
  • Speed to Market

Solution

LegalOn Technologies treated model selection and budget control as part of its engineering system rather than an after-the-fact finance task. After initially giving developers broad access to GPT-5.5 Fast mode and embedding Codex into design, implementation, and daily work, the company saw that unlimited use of top-tier models would exceed budget. Its AI-powered Development CoE created model-selection guidelines backed by testing and monitoring, then paired those rules with administrator controls and monthly usage limits for departments and individuals. Under the new operating model, GPT-6 Luna handles everyday implementation and simpler delegated tasks, GPT-6.1 Sol supports standard design, analysis, and documents, and GPT-6 Astra is reserved for complex architecture and orchestration. Budgets are also tuned to business maturity, keeping high-growth areas aggressive while requiring established units to improve efficiency without slowing overall delivery.

Data flow

Engineering tasks are routed through Codex and then assigned to Luna, Sol, or Astra depending on complexity and delivery stage. Usage data is monitored through administrator settings and budget caps so teams can keep shipping while leadership reallocates AI spend across mature and growth businesses.

Solution architecture

4 components · 2 layers
  1. Compute
    • GPT-6 Luna Handles lightweight implementation, automations, and simpler delegated engineering tasks.
    • GPT-6.1 Sol Supports standard software design, analysis, and document creation that require moderate reasoning.
    • GPT-6 Astra Handles complex analysis, architectural design, and orchestration for the hardest engineering work.
  2. Orchestration
    • Codex Provides the AI coding surface embedded across LegalOn’s development process and daily engineering work.

Architecture clues

  • AID CoE tests and monitors models, then distributes model-selection guidance to teams
  • Administrator settings enforce monthly usage limits at department and individual levels
  • Company is building a feature-release level ROI pipeline to connect AI costs to customer value
  • Fast mode is restricted by default and enabled only when needed

Evidence from the source

"GPT‑6 Astra ... handles advanced judgment, including architecture design"
"GPT‑6 Luna ... handles code implementation, mainly as a subagent"
"Together, model selection, feature restrictions, and budgets tailored to each business reduced estimated daily costs by approximately 65%."
"the company halved costs while maintaining development speed"
Transparency Evidence & methodology 95/100 reference value · 89% confidence · 1 source

Review how this reference was scored and trace each normalized claim back to its documented source.

01

Scoring methodology

How scores are calculated

Customer Reference Value measures usefulness as a customer reference. Evidence Confidence measures the quality and traceability of available evidence, not independent verification.

Customer reference value95/100
  • Problem clarity20/20
    Measures whether the customer's operational problem is specific enough to compare.
  • Implementation depth25/25
    Rewards named components, data flow, integrations, and architecture detail.
  • Outcome specificity20/20
    Measures whether the reported benefit is concrete rather than a generic claim.
  • Quantified impact25/25
    2 metrics describe measurable impact.
  • Deployment scale5/10
    Rewards evidence of a live or scaled deployment and meaningful integration scope.
Evidence confidence89%
  • Source authority20/25
    The main evidence is an official vendor-published customer story.
  • Claim traceability21/25
    10 of 12 key fields link to a named source.
  • Direct evidence20/20
    4 direct source quotes retained.
  • Date provenance10/10
    The publication date comes from the source page or publisher.
  • Source transparency8/10
    1 source reference document publisher, URL, authority, retrieval, and supported fields.
  • Implementation specificity10/10
    Rewards source-backed components, data flow, quantified impact, and a specific outcome.
02

Traceability

Sources & provenance

Field coverage shows which facts each source supports. Dates are shown only when documented by the source.

  1. OpenAI

    LegalOn Technologies customer story

    First-party
    https://openai.com/index/legalon-halves-codex-costs/
    Source type
    Vendor-published customer story
    Published
    Sep 23, 2026
    Supports
    • Customer
    • Business problem
    • Solution
    • Products
    • Solution components
    • Data flow
    • Integrations
    • technical area
    • use case category
    • Business outcome
    • outcome category
    • Metrics
    • Architecture
    • evidence quotes
    • Publication date

Date evidence

  • Published on source page Sep 23, 2026 · OpenAI · First-party
    Open date source
Not separately sourced Industry, Region, Company size. Treat these normalized attributes as lower-confidence until a dedicated source is added.

Related

Similar use cases

See all like this
OpenAI Oct 9, 2026 Asana Software / SaaS North America Asana needed to cut the cost and latency of its StackAI browser agent so it could offer customers more capable models for browser-driven workflows without unsustainable operating expense. $0.47 average estimated model cost with the optimized GPT-6.1 Sol workflow Application RefactoringIT Operations Automation Codex · Command · GPT-6 Astra OpenAI Oct 9, 2026 Sophos Software / SaaS Global Sophos needed to investigate and respond to a growing volume of cyber threats faster, while scaling its expertise across every protected customer without matching growth in scarce analyst headcount. 52% of MDR cases resolved end-to-end by AI IT Operations AutomationSecurity and Compliance OpenAI Daybreak · OpenAI models · Sophos Fusion OpenAI Oct 8, 2026 Braintrust Software / SaaS North America Braintrust wanted to respond to customer feature requests faster and run more engineering experiments, but slower coding tools and backlog-driven workflows made it costly to turn feedback into working prototypes in real… 50% of the Braintrust team moved to Codex in one month Application RefactoringIT Operations Automation Codex · GPT-5.5