Blogs
Practical AI: How Scrubbed Professionals Built Solutions for Everyday Bottlenecks

At A GlanceDuring the inaugural Scrubbed Workflow Hackathon, 66 professionals built practical, secure AI prototypes to resolve the exact operational friction points they face daily. By equipping our teams to systematically address routine bottlenecks, we reduce manual rework and preserve critical bandwidth.On July 15, 2026, 66 Scrubbed professionals gathered on-site at Scrubbed Tech Center for the inaugural Scrubbed Workflow Hackathon. Led by Kendrick Kho, Scrubbed’s Chief AI Officer, the one-day internal learning workshop challenged teams to step away from their standard routines and apply artificial intelligence to the daily friction points they encounter.The objective was highly practical: identify a recurring, time-consuming process or task, and build a working AI solution to fix it.To ensure these proof-of-concept solutions were viable for actual finance operations, teams were evaluated against a rigorous 25-point scoring framework. Judges assessed the prototypes not just on technical capability, but on their ability to solve real workflow pain points and deliver measurable business value. The evaluation placed significant weight on practical implementation, including workflow completeness, scalability, and strict data security guardrails. Ultimately, the strongest innovations were those grounded in real, everyday challenges and designed by people who understand these workflows firsthand.Here is a closer look at the winning prototypes and the teams behind them.1st Place: FSG Connect HubTeam 3 – Functional Support Group (FSG) Members: Christian Ponce, Carmela Therese Lagman, Ruth Angela Dela Cruz, David Justin Tinio, Catherine Louise VitugMentor: Dovie MayoresThe ChallengeNavigating internal support often involves guesswork. Employees frequently know they need assistance, whether it involves HR, payroll, IT support, or marketing. But they often spend unnecessary time figuring out exactly which department handles the request, leading to multiple redirects and delays.The SolutionTeam 3 developed the FSG Connect Hub, an AI-powered internal directory and service portal.Instead of requiring employees to memorize organizational structures, the hub was designed to allow users to ask questions naturally. The AI engine guides them directly to the correct team, service, or resource. By consolidating a service catalog and knowledge base into a single, intuitive interface, the tool demonstrated how to eliminate the typical endless loop of internal inquiries.When our professionals spend less time navigating internal administrative hurdles, they preserve their bandwidth for what matters most: accelerating financial execution and maintaining steady, reliable client delivery. Key Takeaway from the Team"Innovation does not always have to come from solving highly complex problems. Sometimes the greatest impact comes from improving everyday experiences and removing friction from common processes."2nd Place: Rec 'N' RollTeam 10 – Technical Consultancy and Client Accounting Services (TCCAS) Members: Dexter Ayson, Kimberly Claire Manalili, Christian Dhey Usi, Angel Lou RuizMentor: Kenji CanapiThe ChallengeCash reconciliation is a universal pain point in the accounting industry. It requires highly manual, line-by-line checks to match bank transactions with ledger entries, which drains time and increases the likelihood of human error.The SolutionTeam 10 built Rec 'N' Roll, an AI-powered dual-ledger credit card and bank reconciliation tool.The prototype leverages AI to cross-reference bank feeds with internal records in seconds, effectively handling date mismatches, transposition errors, and fuzzy description matching. To combat task fatigue, the team integrated user-experience features: a "Rec 'N' Roll" mode, a "Cheer Me Up" button, and a mini-game designed to help users maintain focus and accuracy during high-volume reconciliation periods.Key Takeaway from the Team"Workflow automation isn't just about moving data from point A to point B; it's about giving the user the tools, the clear insights, and the mental bandwidth... to make better decisions, faster."3rd Place: Financial Package Generator & InvoiceIQBecause the competition was tight, two teams secured 3rd place by tackling different, yet equally critical, data entry bottlenecks.(Note: Because both of these workflows involve parsing sensitive financial information like Trial Balances and invoices, all AI testing and prototyping were conducted within strict, secure data guardrails to protect client confidentiality and ensure safe-use compliance.)Team 6 – Sector Accounting Team 1 (Financial Package Generator)Members: Victorina Leonila Mae Santos, Stephanie Nicolle Vital, Karla Joy Rombaoa, Trisha Nicole MagcalasMentor: John Christian DonatoThe ChallengePreparing monthly management reports is notoriously tedious. Exporting massive Trial Balances (TB) from various accounting systems and manually structuring the data in Excel frequently causes system lag, stretching a single client report preparation to 3–4 hours.The SolutionThe team developed the Financial Package Generator, a two-step AI workflow designed to eliminate manual data formatting entirely.The tool automatically digests raw TB exports and maps them into standardized P&L reporting formats. Furthermore, it utilizes AI to identify Month-over-Month (MoM) variances and to draft preliminary commentary on revenue shifts and expense spikes. This allows preparers to shift their focus from raw data assembly to strategic review and margin evaluation.Key Takeaway from the Team"AI is an accelerator, not a replacement for domain expertise. AI provided the engine to automate data parsing and initial drafting, but our accounting background was vital to guide the prompts, set the logical rules, and validate the output."Team 13 – Sector Accounting Team 5 (InvoiceIQ)Members: Aimee Marie Lugtu, Thea Louise Mirasol, Hercel Mae Muñoz, Rachelle Ann RomeroMentor: Kris SantiagoThe ChallengeFinance teams spend countless hours manually reading PDF invoices, extracting dates and amounts, and looking up General Ledger (GL) codes. Traditional OCR systems frequently struggle with complex layouts and contextual mapping. The SolutionThe team mocked up InvoiceIQ, an AI-powered document parsing and bookkeeping tool. Instead of just reading text, the AI was programmed to understand the business purpose behind each line item, extracting verbatim descriptions and intelligently mapping them directly to a company’s Chart of Accounts to generate structured, Google Sheets–ready CSV reports. Key Takeaway from the Team "We learned that clear problem definition and precise instructions are far more critical than complex code. Building effective AI solutions is about acting as an architect—giving the model unambiguous layout constraints, strict field boundaries, and concise context yields far better results than trying to over-engineer a complex system."Looking ForwardThe first Scrubbed AI Workflow Hackathon proved a vital operational theory for modern businesses: the most effective technological advancements are those driven by user-centric design. For Kendrick Kho, who led the event, it was a test of what happens when finance teams are asked to cultivate brand-new technical skills with minimal hands-on coaching. The results far exceeded expectations, successfully hitting three major operational outcomes: producing at least one prototype ready for official development by the tech team, proving to Scrubbed leadership that this training yields tangible ROI, and sparking a new culture of building among the staff.So, what comes next?The vision for the Hackathon program is to expand both its reach and complexity. Across the organization, the goal is to cultivate "citizen developers" at every level and within every group. While this inaugural event focused on the fundamentals of building for non-coders, future iterations will shift toward "Building AI for Non-Coders"—incorporating more advanced technologies into exactly what our teams are capable of creating.By equipping professionals with AI tools to solve the exact bottlenecks they face, Scrubbed aims to drastically reduce manual labor, minimize errors, and create more bandwidth for high-level financial analysis.Key Takeaways:When finance starts consuming more time than it should, it is usually a sign the execution layer needs support, which our professionals addressed by proactively building AI prototypes to resolve their exact daily bottlenecks.Because strong execution should reduce pressure rather than add to it, all AI testing was conducted within strict data guardrails to protect client confidentiality and maintain operational rigor.Recognizing that data is only as reliable as the system and team that generates it, these AI accelerators rely entirely on our professionals' domain expertise to guide the rules and validate the outputs.By systematically automating routine reconciliation and internal routing, our teams reduce manual rework and give leaders back time, focus, and confidence.