SCSK has announced the CELF AI Development Agent, a no-code app builder powered by Anthropic's Claude Code via Amazon Bedrock, targeting the 82.6% of Fit-to-Standard ERP adopters that still rely on spreadsheets for peripheral workflows [1]. Alpha testing with four partner firms delivered implementation productivity gains of 300–800% per screen versus conventional development [1]. The beta launch on October 23, 2026 arrives as 86.7% of AI consulting sellers expect that service to drive growth in 2026 [2], signaling strong channel tailwinds for SCSK's productized approach.
What is Covered in this Article
- Structural demand: why 82.6% of ERP adopters still run peripheral workflows in Excel [1]
- Product mechanics: how Claude Code and Amazon Bedrock power CELF's agentic development loop [1]
- Alpha validation: 300–800% productivity gains across four partner firms [1]
- Channel conviction: 86.7% of AI consulting sellers see growth ahead, with the AI consulting market on a strong growth trajectory [2][3]
- Roadmap: citizen-developer version planned for spring 2027 [1]
The News: SCSK announced on October 14, 2026 that it will launch the beta of its 'CELF AI Development Agent (provisional name)' on October 23, 2026, recruiting up to 30 companies through November 20 [1][1]. The service uses Anthropic's Claude Code, accessed via an Amazon Bedrock environment provided by SCSK, to generate CELF app screens, relational database tables, and business logic from natural-language chat instructions [1]. The CELF AI Development Agent component is free in beta; Claude Code usage fees are charged separately [1]. Alpha testing with CELF Partner Network members showed implementation productivity gains of 300–800% versus conventional development across four partner firms [1]. SCSK will showcase the service at CELF CONNECT 2026 in Shinagawa, Tokyo in November 2026 [1].
SCSK Bets on Agentic AI to Digitize Japan's Excel-Dependent Enterprises
Analyst Take: SCSK is addressing a well-documented structural gap in enterprise digitization: the persistent reliance on spreadsheets for workflows that standard ERP systems cannot absorb. By embedding agentic AI directly into its no-code platform, SCSK is not merely adding a feature but repositioning CELF as the primary vehicle for what it calls 'AI-ready' transformation of Excel-based operations. The early productivity data from alpha testing is compelling, and the channel dynamics are favorable.
A Large, Underserved Market Hiding in Plain Sight
SCSK's own survey of 1,032 respondents found that 82.6% of enterprises pursuing Fit-to-Standard ERP adoption still have business areas that standard functions cannot cover, with spreadsheet software being the most common workaround [1]. This is not a niche problem. Approximately 90% of CELF apps already in production are concentrated in ERP-peripheral areas such as budget and actual management and deal management [1], confirming that the market has been self-selecting into exactly this gap. The commercial signal is equally clear: CELF-related ERP peripheral business cases grew approximately 3.4 times year-over-year comparing FY2025 to FY2026 through mid-August [1]. That growth rate suggests demand is accelerating faster than conventional no-code tooling can serve it, which is precisely the opening SCSK is exploiting with an agentic layer.
Agentic Architecture: Claude Code Inside a No-Code Shell
The CELF AI Development Agent uses Anthropic's Claude Code via an Amazon Bedrock environment provided by SCSK, allowing developers to describe requirements in natural language and have the agent generate screens, relational database tables, and business logic within CELF's visual programming environment [1]. A key design decision is transparency: AI-generated logic is rendered in CELF's native visual programming interface, preventing the black-box problem that often undermines enterprise adoption of AI-generated code. Alpha testing with four CPN partner firms, covering apps ranging from vegetable processing records to tenant security deposit management, showed implementation productivity gains of 300–800% versus conventional development on a per-screen basis [1]. S'PLANT reduced a single screen from 16 hours to 2 hours; NSK cut a procurement management screen from 48 hours to 7 hours. These are not marginal improvements.
Channel Timing: Riding the AI Consulting Wave
The beta launch lands at a moment of peak channel conviction around AI services. The Futurum Ecosystems, Channels and Marketplaces Decision Maker Survey, 2H 2026 found that 86.7% of AI consulting sellers expect AI consulting to drive growth for their business in 2026 [2], and 78.3% of AI software sellers expect AI software, including copilots, to drive growth in 2026 [2]. Critically, 66.8% of partners confident of succeeding in an AI-transformed market have developed their own AI solutions using LLMs [2], consistent with SCSK's strategy of embedding Claude Code into its own platform rather than reselling a generic tool. Custom application development also remains high-conviction: 71% of application development sellers expect that service to drive growth in 2026 [2]. The broader market context reinforces the opportunity, with the Channel Ecosystems market base forecast at $25,680M in 2026, growing at a 36% CAGR from 2022 to 2029 [3]. SCSK's move to productize AI-assisted development services positions it squarely at the intersection of the two highest-conviction growth vectors its channel peers are pursuing.
Roadmap and Ecosystem Reach
SCSK's near-term roadmap extends the agentic capability beyond IT developers. A citizen-developer-oriented version is planned for spring 2027, targeting business-unit users who can independently build apps for their own Excel-based workflows [1]. This two-tier model, where partners and IT departments handle complex upstream work while business users self-serve simpler automations, mirrors the deployment patterns emerging across the broader no-code and low-code market. CELF's existing installed base of more than 1,500 companies globally and more than 130 partner firms in its CELF Partner Network [1] provides a ready distribution channel for both the beta and the eventual commercial release. The free-component beta structure, capped at 30 companies, is a deliberate land-and-expand motion designed to generate reference cases before pricing is announced.
What to Watch
- Beta conversion rate: how many of the 30 beta participants proceed to paid deployments once commercial pricing is announced after November 20 [1]
- Productivity benchmarks at scale: whether the 300–800% per-screen gains from alpha testing [1] hold across more complex, multi-screen enterprise apps in the beta cohort
- Citizen-developer launch readiness: whether SCSK's spring 2027 target for the business-unit-facing version [1] holds given the complexity of abstracting further from the developer workflow
- CPN partner adoption: how quickly the 130-plus CELF Partner Network firms [1] embed the agent into their own service offerings, which would be the primary signal of channel-level conviction
- Competitive response: whether rival Japanese IT service providers or global no-code vendors introduce comparable LLM-native development agents targeting the same ERP-peripheral segment
Sources
1. CELF、Excel業務のAI-ready化を支援する「AI開発エージェント」を提供～生成AIとの対話で業務アプリ開発を効率化、β版参加企業を募集～, Scsk
2. 2H 2026 Ecosystems, Channels & Marketplaces Global Enterprise Decision Maker Survey Report, Futurum Research, August 2026
3. 2H 2025 Hyperscaler Marketplace Market Sizing & Five-Year Forecast, Futurum Research, December 2025
Disclosure: Futurum is a research and advisory firm that engages or has engaged in research, analysis, and advisory services with many technology companies, including those mentioned in this article. The author does not hold any equity positions with any company mentioned in this article.
Read the full Futurum Group Disclosure.
Other Insights from Futurum:
SCSK and Vector Bet on Edge ECU Co-Creation to Win SDV Software
SCSK Bets on Vertical AI Agents to Crack Japan's Real Estate Market
SCSK Embeds Generative AI Into Workforce DNA
Author Information
This content is written by a commercial general-purpose language model (LLM) along with the Futurum Intelligence Platform, and has not been curated or reviewed by editors. Due to the inherent limitations in using AI tools, please consider the probability of error. The accuracy, completeness, or timeliness of this content cannot be guaranteed. It is generated on the date indicated at the top of the page, based on the content available, and it may be automatically updated as new content becomes available. The content does not consider any other information or perform any independent analysis.
