An eight-Saturday course in Colombo on working with AI, including where it is useful and where it should not be trusted.
Excl. VAT & SSCL
None of this is about being more rigorous than you already are. It is about knowing which parts of the job these tools can safely take on, which comes from using them rather than reading about them.
You have been asked to look at AI, and you want to know what happens when it gets a calculation wrong.
You are being asked to approve other departments' AI subscriptions without much idea what they actually do.
At month end you pull figures out of several different systems into one spreadsheet by hand, and it usually takes up a weekend.
Once the figures are in, writing the commentary that explains why they moved takes longer than assembling them did.
Matching invoices against purchase orders and bank statements takes up a large part of your team's week.
Every session produces something you keep. These are the things every participant leaves with.
Text, voice, image and code. In a finance role that mostly means handling documents rather than doing calculations.
Claude Code, Obsidian, Wispr Flow and Motion set up around your own work, plus access to a library of more than 1,000 AI tools across 40 categories.
You are often the person approving software you had no part in choosing. You learn how to assess one on its own terms.
Including an assistant trained on your own policies and past filings, and an automation running on real work.
Week 6 goes deeper on the work closest to what you do, with someone who does that work using these tools.
A written plan for the three months after the course, reviewed by the rest of the group, that you can show to whoever needs to approve the work.
A private network connecting BuildrLabs graduates with companies, founders and professionals across the Hatch group.
Graduates keep access, maintained and updated continuously.
Explore the libraryEvery session produces something you keep. Seven Saturdays run in person at Hatch Works; Week 7 is an online build day.
Claude Code, Obsidian, Wispr Flow and Motion, set up and working by the end of the day
Four modules. One is a review of your own working week, which for most finance people shows how much time goes on assembling reports and answering the same requests.
An assistant trained on your company documents, which you can ask questions
Your policies, past filings and prior-year treatment become searchable. This includes reading documents that only exist as scans.
A 60-second video advert with script, voiceover, visuals and music
Image, video, audio and voice generation using Midjourney, Ideogram, Runway, ElevenLabs and Suno.
A working web application on a public address
Three modules covering automation in Zapier and Make, agents that run without supervision, and building an application. Also covers APIs, which is how these tools connect to the systems you already use.
A 90-day plan and a written AI usage policy for your team
What data can and cannot go into a tool, how to work securely, hallucination and accountability, and how GDPR, the AI Act and industry rules apply.
An assistant that answers a real question from a colleague, and an automation your team would use
A full day going deeper on the work closest to your own. You work out where AI would help most in your role, build an assistant on your own documents, then build an automation and test it against real data.
Your assistant, automation and application combined into one thing
A full day of building with facilitators available, plus a rehearsal of your demo.
Your finished project, presented
You demonstrate what you built to the rest of the group, get feedback, and receive your certificate.
The November 2026 cohort brings marketers, sales leads, finance teams, HR heads and business owners into the same room at Hatch Works. For five Saturdays you build alongside people who do different jobs to you, so you see how finance automates month end while someone else automates a campaign.
Operations and logistics people join too, with their own Week 6 track.
If you use a laptop for work, you are qualified.
Excl. VAT & SSCL
Eight Saturdays at Hatch Works, Colombo, every session, materials, tool access and your certificate. An instalment plan is available: ask us and we confirm it in writing before you pay.
Everyone teaching on this course does the work for a living. The group includes heads of AI and founders, along with digital and transformation leads from Dialog Axiata, MAS KREEDA, John Keells Group and Wavenet. In Week 6 you go deeper on your own function, with someone who works in it.
More trainers join each cohort as new domain tracks open.
Learn from them. Enrol nowHallucination and accountability are taught directly in Week 5, so you learn how to design around the problem rather than hope it does not happen. Week 5 also covers what data you can put into a tool when you work in a regulated environment.
For the work that surrounds the numbers rather than the calculations themselves: drafting report commentary, reading and routing documents, summarising what changed between periods, setting up forecasts that can be rerun with new assumptions, and assembling compliance paperwork. Most AI for finance work sits in that surrounding layer.
No. It can take on repetitive work that follows clear rules, such as routing documents, producing first drafts and summarising. Judgement, sign-off and professional responsibility stay with a person.
Only if you set the conditions yourself. Week 5 covers what should and should not be shared with a tool, how to work securely inside a company, and how GDPR, the AI Act and industry rules apply. Some of the workflows taught do not require you to share source data at all.
The four you set up in Week 1 and use throughout are Claude Code, Obsidian, Wispr Flow and Motion. Alongside those you work with Claude, ChatGPT and Gemini for text, and Zapier, Make and n8n for automation.
Week 4 covers APIs and connectors through Zapier, Make and n8n, which reach most common finance systems. Your Week 6 project is scoped against the systems you actually use, so bring those details with you.
Yes. There is no prerequisite, and where something needs code, Claude Code writes it while you tell it what you want. Most of building an automation is working out the rules, which is already part of your job.
Bring whatever you are comfortable bringing. Week 5 covers the boundaries in detail, and everything you build will work on sample or redacted material.