ForgeVision AI Fast Delivery Service

Overview Traditional development defines every requirement up front and builds for months or years - by go-live, the business context has changed. ForgeVision's AI Fast Delivery Service fixes the timeline, team, and budget first, then builds, shows, and refines in short cycles. We follow AIDLC (AI-Driven Development Life Cycle), the AI-driven development methodology introduced by AWS in 2025: AI handles implementation while our engineers focus on judgment and review. Because AI produces design, code, and tests at the same time, the first working version takes roughly one-fifth to one-tenth of the conventional effort.

What you get with Fast Delivery (2 weeks, fully remote)

• A working prototype application (PoC): 1-2 screens and 2-5 functions (APIs)
• The complete source code
• AI-generated design documentation
• Goal: confirm the direction of the solution on a working screen
• Your time commitment: 3 meetings, 4 hours in total (120-minute discovery, 60-minute mid-point review, 60-minute final demo)
• AI tools used: Claude Code (Anthropic) or Kiro (AWS) If the requirements exceed what can be built in 2 weeks, we narrow the scope together before starting.

How the 2 weeks work (10 business days)

• Day 0: Discovery session (120 minutes, free) - we start from this session alone
• Day 1-2: Organize your goals with AI and fix the scope
• Day 3-5: AI drafts screen and feature designs; our engineers review them from a business point of view. Mid-point review (60 minutes) on Day 5
• Day 6-8: AI generates code and tests together and assembles the working application
• Day 9-10: Quality and security checks, final demo (60 minutes), and hand-over of all deliverables

After Fast Delivery: decide whether to continue Review the deliverables, then decide. If you continue, choose one of three menus. The monthly fee is set by the team size and the duration by the menu, so your budget (monthly fee x months) is fixed before work starts.

• Greenfield (build new): new development + ongoing operations support. Small: from 3 months / Medium: from 6 months / Large: from 12 months
• Brownfield (rebuild existing): AI reverse engineering of your existing system (1-3 months) + rebuild (3-12 months) + ongoing operations support
• Continuous (improve while running): AI reverse engineering (1-3 months), then continuous improvement on the monthly subscription AI reverse engineering recovers up-to-date specifications from the source code of an existing system using AI.

Monthly subscription (enhancements and new features included) A fixed monthly fee for a dedicated team. Enhancement requests are handled within the monthly capacity - no quotation negotiation for each change, so IT budget becomes a fixed cost.

• Small: 1 AIDLC engineer, up to 80 hours/month, approx. 2-3 enhancements per month
• Medium: PM + 2 AIDLC engineers, up to 170 hours/month, approx. 3-5 enhancements per month
• Large: PM + AIDLC coach + 3 AIDLC engineers, up to 260 hours/month, approx. 5-8 enhancements per month Work is delivered on a best-effort basis within the contracted capacity; priorities are agreed monthly, and requests beyond capacity are carried over or quoted separately. Because AI keeps the design documents up to date, each enhancement typically takes 1-5 business days.

Compared with conventional SI

• Time to start: 3-6 months of proposal and approval -> working software 2 weeks after discovery
• Enhancement lead time: weeks to months -> 1-5 business days per change
• Billing for changes: quoted every time -> included in the monthly fee
• Design documents: stale after release -> kept current by AI
• Team continuity: separate build and run teams -> one team for 1-2 years

Customer example (manufacturing company)

• Enhancement lead time: from 2-3 months to an average of 2 weeks
• Enhancements per year: from 3-4 to 12-24
• Design documents: from 5 years out of date to always current
• Budgeting: from per-change quotations to a fixed monthly fee

Why ForgeVision

• AWS Advanced Tier Services Partner and Anthropic (Claude) reseller for Amazon Bedrock
• ISO/IEC 27001 (ISMS) certified since March 2024
• Secure use of AI: your information is never used to train AI models; personal data is masked
• No single point of failure: a primary and a secondary engineer on every engagement, design documents always current, capacity confirmed before we accept

Pricing and how to buy All engagements are sold through AWS Marketplace private offers and billed on your AWS invoice. Choose "Request private offer" or contact us to schedule the free 120-minute discovery session; we reply within 2 business days. Prices are exclusive of Japanese consumption tax. If the prototype or application is deployed into your own AWS account, AWS infrastructure charges incurred there are separate from the AWS Marketplace transaction and are your responsibility.

Delivery: fully remote. Language: Japanese (contact us regarding English). Region: Japan.