Thirty years from overnight batch to executive officer

Mandate

Make the firm's dependence on a shared back-office service governable, and keep every channel usable for an older customer base while branches close.

Outcomes

• Set a partial order-acceptance rule during a supplier batch failure; no settlement failed
• Made passkeys compulsory for 41,000 online accounts from October 2025
• Cut median inheritance transfer time from 94 days to 41
• Supported the reduction from 34 branches to 27 with tablets and video consultation booths

Decisions in context

The board created this role when it separated technology from general administration, and I was its first holder. Most of my time goes on three relationships. The first is with Kuranoe, whose service runs our overnight processing and whose failures become ours the moment a branch opens. The second is with sales and compliance, because every incident rule and security control changes what an adviser can offer a customer. The third is with the board, which needs system risk described in customer terms rather than server terms. The three cases on this site come from this appointment, and each was approved above me. I hold no directorship. I attend the board to present and answer questions, and I report separately to our audit and supervisory board members, who sometimes ask harder questions than the directors do.

Reason for the move

Current appointment; considering one outside-director seat with the president's consent.

Mandate

Turn the first years on the shared service into a managed relationship and modernise the tools our advisers used.

Outcomes

• Agreed a service-level schedule with Kuranoe that included customer-facing deadlines
• Issued tablets to 560 sales staff in 2019
• Kept home-visit and branch service running through 2020 with remote statement printing and telephone consultations

Decisions in context

After the migration, the firm needed someone to manage what it had bought. Kuranoe's standard service levels described when its batch finished, not when our customers could see a correct statement. I spent most of 2017 rewriting the schedule with Kuranoe's account team so that each deadline was tied to something a customer or branch needed, and so that a missed deadline triggered a named call. In 2019 we gave tablets to all 560 sales staff, replacing paper portfolio summaries that advisers printed before home visits. The pandemic tested that work in 2020, when many older customers would not come to branches. We printed statements centrally, moved most consultations to telephone and kept order-taking open. My 22 colleagues carried that difficult year, and it convinced the board that technology needed its own executive voice.

Reason for the move

Appointed the firm's first executive officer for IT when the board separated technology from general administration.

Mandate

Move the back office from the firm's own mainframe to a shared service before the mainframe maintenance contract ended in 2016.

Outcomes

• Cut over during Golden Week 2016 without invoking the rehearsed rollback
• Reduced 214 bespoke reports and processes to 61
• Retained the firm's own daily copy of executed orders and cash movements

Decisions in context

I was hired to do for Ogiwatari what I had done for clients in Osaka, and to avoid the mistake I had made there. We tested Kuranoe's processing thoroughly, but we spent as long testing our own end-of-day habits: which lists branches printed, which reconciliations operations ran by hand, which reports compliance read each morning. Of 214 bespoke reports and processes, we kept 61. Each removal was agreed with the person who used the report, after they had explained the decision it supported. I also insisted that we keep our own daily copy of executed orders and cash movements outside the shared service. Some colleagues saw that as needless duplication. It cost about ¥30 million a year and proved its worth eight years later. The cutover ran over Golden Week in May 2016, with two full rehearsals beforehand.

Reason for the move

Promoted to general manager of IT planning to run the supplier relationship the migration had created.

Mandate

Deliver system conversions for regional securities firms, shinkin banks and leasing companies.

Outcomes

• Led two securities back-office conversions to packaged and shared services
• Introduced end-of-day reconciliation testing built from clients' own branch routines
• Delivered a shinkin bank's teller-terminal replacement across 41 branches

Decisions in context

Yodomaru gave me my first experience of being the supplier, which has made me a more realistic customer. As a systems engineer I wrote conversion programs and test plans; as a project manager from 2009 I answered to clients' steering committees for scope, cost and dates. The two securities conversions I led both went live on schedule. The second left its client with a month of manual reconciliations, because our testing had followed the new system's design rather than the way the client's branches actually closed their day. I rebuilt our test approach afterwards around clients' own routines, and the shinkin bank terminal project that followed went in cleanly across 41 branches. I learnt that a supplier can meet every contractual milestone and still leave the client's own staff carrying the risk after hours, unpaid and unseen.

Reason for the move

Recruited by Ogiwatari to lead its own back-office migration in Takamatsu.

Mandate

Keep card authorisation dependable while merchants moved to chip terminals.

Outcomes

• Specified offline limits and fallback behaviour for 4,800 merchant terminals
• Reduced authorisation timeouts during evening peak hours
• Documented which losses fell on the company when offline approvals were later declined

Decisions in context

Card authorisation taught me the value of a fast, honest answer. A shopper at a supermarket till in Niihama cannot wait for an overnight batch, and a terminal that hangs is worse than one that declines. My work was to specify how 4,800 merchant terminals behaved when our host or the network was unavailable: which purchases could be approved offline, up to what limit, and how those approvals were reconciled later. Each limit was a trade between customer convenience and credit loss, and I learnt to put that trade to the credit and merchant managers in plain figures rather than settle it as a technical parameter. I also worked on the evening peak, when timeouts clustered. I left after five years because I wanted to see how other financial institutions made the same decisions.

Reason for the move

Moved to Osaka for wider project responsibility across financial institutions.

Mandate

Maintain COBOL batch programs for credit cooperatives and municipal payrolls.

Outcomes

• Maintained deposit and loan interest programs for three credit cooperatives
• Remediated and tested year-2000 date logic with client staff
• Introduced output-total checks before batch results were released each morning

Decisions in context

I joined straight from university with a degree in information science and very little idea of how a cooperative calculated interest. The senior engineer I worked under insisted that every change be checked against the previous night's totals before results went to clients, and that habit has stayed with me for thirty years. Most of my programs ran between midnight and five, and their output reached cooperative counters when doors opened. From 1998 the year-2000 work took over: reading old date logic, correcting it and testing it with clerks who knew what a wrong figure looked like long before I did. Nothing failed on the first working day of 2000. I left that autumn wanting to work on systems that responded to a customer immediately rather than overnight, which took me across the prefectural border to Matsuyama.

Reason for the move

Joined a card company to work on systems that answered customers in seconds rather than overnight.