0
30 March 2026 9:30:00 AM
11 min
When the Room Agrees: AI is Reshaping Finance — But Only If Your Foundation is Ready There are conferences where you hear ideas you've never encountered before. And there are conferences where the room puts language to things you've been seeing in practice — where the conversations happening on stage mirror almost exactly what you've been building with clients behind the scenes. Today's Gartner ...
When the Room Agrees:
AI is Reshaping Finance — But Only If Your Foundation is Ready
There are conferences where you hear ideas you've never encountered before. And there are conferences where the room puts language to things you've been seeing in practice — where the conversations happening on stage mirror almost exactly what you've been building with clients behind the scenes.
Today's Gartner Finance & Accounting Symposium in Sydney was firmly the latter. I attended a standout panel session that brought together three of the sharpest finance leaders in the room: Matt Clifford, GM Finance Systems & Data at News Corp Australia, Winny Puthussery, Head of Group FP&A at Metcash, and moderated with precision by Jason Codespoti, CFO at IBM Australia.
The topic was AI and the future of finance. But the conversation that unfolded went far deeper — into culture, data architecture, transformation sequencing, and what the finance function actually looks like in 2030. What made it particularly resonant for us at Octane: all three organisations on stage are running IBM Planning Analytics (TM1) as their planning and forecasting platform.
This post captures the key themes from the session — and connects them to the work we've been doing with News Corp on their own Planning Analytics modernisation journey.
"It wasn't the technology that held us back. It was our data. We had to go back and rearchitect from the data layer up."
— Matt Clifford, GM Finance Systems & Data, News Corp Australia · Gartner Finance Symposium 2026
Winny opened with a line that drew knowing laughter across the room: he'd already run 15 budget scenarios that week. It was Tuesday.
It landed because every finance leader in that auditorium recognised the feeling. The panel's first theme was clear: the nature of financial planning has fundamentally changed. The question finance teams are now being asked is no longer "what happened last month?" It's "what happens if I change X — and simultaneously, what happens if I change Y?"
This shift from retrospective reporting to real-time, forward-looking scenario analysis isn't a trend — it's the new baseline. And it has enormous implications for both the tools finance teams use and the processes they build around them.
The panel cut through the AI hype quickly. Winny framed it well: there are two distinct buckets of AI value in finance right now, and conflating them is a source of a lot of frustration.
The Two Lanes of AI Value in Finance
Lane 2 — Foundation Required: The larger prize — connected forecasting, automated variance analysis, AI-driven scenario modelling — but only accessible once your data is structured, governed, and connected across systems. This is where the hard yards are. And this is where most organisations are still working.
Matt's point on Lane 2 was the most candid moment of the session. News Corp had early, ambitious AI use cases ready to deploy — but when they tried to turn them on at scale, the platform wasn't what held them back. It was the data. The structure, the shape, and the care given to that data over the years hadn't been sufficient to make it consumable by AI. The response: a deliberate rearchitecting of the data layer from the ground up.
This is something we see consistently across enterprise planning environments. The TM1 platform is rarely the constraint. It's the state of the data beneath it.
Matt's contribution to this panel carried real weight because it wasn't theoretical. News Corp has been on a live modernisation journey — and Octane has been privileged to be part of it.
The challenge News Corp faced isn't uncommon. A large, complex media organisation — multiple mastheads, traditional print alongside digital, broadcast assets — each with their own planning models, data structures, and forecasting rhythms. The legacy TM1 environment had served them, but it had grown organically over time. Connecting it, governing it, and making it ready for the next generation of AI-driven finance capability required a deliberate platform investment.
35–40 Hours saved per reporting cycle post-modernisation
5–10× Force multiplier on FP&A team capacity with AI orchestration
3 → 1 Planning platforms consolidated onto IBM PA
What the modernisation unlocked wasn't just faster reporting — it was the ability to shift the finance team's focus. Instead of spending the majority of their time moving and formatting data, analysts could direct their energy toward the strategic analysis and scenario modelling that actually drives business decisions.
As Matt described on stage, the goal for News Corp's planning function isn't just automation — it's enabling one person to coordinate work that previously required many. Agents are doing multiple jobs in parallel. The human role is shifting from data handler to decision architect.
"We now have the opportunity to have one person coordinating, where we used to have many. AI gives us that scale multiplier — but only if the platform beneath it is solid.
—Matt Clifford, GM Finance Systems & Data, News Corp Australia · Gartner Finance Symposium 2026
One of the most striking observations from Winny Puthussery was about a structural shift in how organisations view data. He noted that Metcash recently combined its data and AI roles into a single Chief Data and AI Officer position a clear signal that the two disciplines are inseparable.
The reason is straightforward: AI is only as intelligent as the data it operates on. You can deploy the most sophisticated planning and forecasting AI available, but if your data lives in silos, lacks governance, or hasn't been structured with cross-system connection in mind, the AI will surface that fragility immediately.
For Planning Analytics environments specifically, this means the investment in data architecture, model design, and dimension hygiene that might once have felt like internal IT housekeeping is now a direct enabler of competitive advantage. The organisations getting the most value from their TM1 environments today are the ones that treated data quality as a strategic priority before AI arrived at the door.
The panel was unanimous on the transformation strategy. The days of the two-year ERP-style rollout are over, not just because the pace of change makes them obsolete, but because organisations that wait that long are already 18 months out of date by the time they go live.
Matt's framing was practical and repeatable: identify the highest-value use cases, execute them well, demonstrate ROI, build trust, and roll that credibility into the next project. This incremental approach isn't a compromise; it's the methodology that actually sticks because it creates internal champions at every stage.
The Transformation Sequencing Principles — As Heard on Stage
Pick use cases where the win is visible, and the value is unambiguous — workforce planning, forecast cycle time, reporting automation
Invest in the people who know your business and make them owners of the change — not passengers through it
Frame AI as a force multiplier, not a cost-cutting exercise. If your team can do 5–10× more, the ROI conversation changes entirely
Stabilise before you modernise — a well-governed, performant foundation is the prerequisite for everything that comes next
Don't wait for perfection. The tools are improving month by month. Start with what you can manage, and grow from there
Perhaps the most important conversation of the session wasn't about technology at all. It was about people.
Both Matt and Winny acknowledged that the emotional reality of AI in finance — the nervousness, the questions about job security — can't be glossed over. It has to be addressed directly, because that baseline concern is the barrier to even having the conversation about what's possible.
The reframe that resonated most: right now, many highly skilled finance professionals spend 80% of their time moving data around and 20% doing the strategic work they were trained and hired to do. AI inverts that ratio. And when finance teams start experiencing that inversion — when they find themselves spending their day on scenario analysis, business partnering, and strategic insight rather than data gymnastics — the culture shifts naturally.
Jason closed the session with a line worth repeating: "Treat AI as a partner, not a threat. The people who do that well will define the next era of finance leadership."
The panel's closing question asked each participant to look five years ahead. The picture that emerged was consistent across all three voices.
Finance becomes the decision-making backbone of the organisation. Not the back office. Not the reporting function. The team that spends almost no time on data preparation and almost all their time providing the scenario analysis, risk modelling, and strategic input that underpins major business decisions.
The path to that outcome isn't mysterious. It runs through a modern, well-governed planning platform. Connected data. An AI layer that can operate with trust and explainability. And a team that has been developed to be business partners first, data operators second.
For organisations running IBM Planning Analytics (TM1), the foundation is already there. The question is whether it's been built, maintained, and evolved in a way that makes it ready to carry that future.
We help finance teams modernise their IBM Planning Analytics / TM1 environments and build the data and platform foundations needed to leverage AI at scale. If you'd like to discuss where your environment sits and what the path forward looks like, we'd welcome the conversation.
0
13 April 2026 6:00:00 PM
9 min
“We need a chatbot.”
But here’s the uncomfortable truth, Most AI initiatives don’t fail because of poor models.
They fail because the data behind those models is not trusted.
Before organizations can scale AI, they must solve a more fundamental challenge:
Can your business trust the data driving your decisions?
Many enterprises believe they are AI-ready because they have:
Data warehouses
BI dashboards
Reporting tools
Multiple versions of the same KPI
Manual data reconciliations
Lack of lineage and auditability
Delayed reporting cycles
But when you go one layer deeper, cracks start to appear:
This creates a critical issue:
AI models trained on untrusted data only amplify inaccuracies, at scale.
AI does not create intelligence out of thin air.
It learns patterns from existing enterprise data.
If your data ecosystem has:
Inconsistent definitions
Siloed systems
Weak governance
Conflicting
Non-explainable
Difficult to operationalize
Then your AI outputs will be:
This is why organizations struggle to move beyond pilots.
Leading organizations are now rethinking their approach:
Instead of building pipelines, they are building data products:
This shift is critical because AI systems require contextual, trusted, and reusable data assets.
This is where a unified approach becomes important.
Data Foundation – IBM watsonx.data
A hybrid, open lakehouse architecture that enables:
Access to structured and unstructured data
Cost-efficient scaling
Query performance across distributed sources
Provides:
Foundation models
Prompt tuning capabilities
AI lifecycle management
Explainability
Bias detection
Regulatory compliance
Model monitoring
Reports
Forecasts
Scenario simulations
Business decisions
Ensures:
Explainability
Bias detection
Regulatory compliance
Model monitoring
Transforms AI insights into:
Reports
Forecasts
Scenario simulations
Business decisions
Most organizations stop at insight.
But real value is created when you connect:
Data → Trusted foundation
AI → Intelligence generation
Planning → Scenario evaluation
Execution → Business action
This closed-loop system is what defines a mature AI-driven enterprise.
0
10 April 2026 9:45:01 AM
8 min
Not all financial close platforms are equal. Here's exactly what to look for, what questions to ask, and why most vendors will struggle to answer them.
The financial close software market looks more crowded and more confusing than it did five years ago. Task management tools, RPA platforms, ERP-native modules, AI copilots, and now agentic AI systems — all claiming to solve the same problem, all using similar language to describe very different capabilities.
If you're evaluating options in 2026, you need a way to cut through the noise and ask the right questions. This guide gives you exactly that.
We'll cover the five categories of capability that actually matter, the specific questions to put to vendors, and the red flags that separate a genuinely production-ready platform from a roadmap wrapped in a demo.
Before we get to the evaluation framework, it's worth being clear about why so many finance teams have implemented close software and still spend 8 days closing.
The most common failure mode isn't bad technology — it's a mismatch between what the tool does and what the close actually requires.
Task management tools
These are the most widely deployed categories: checklists, close task boards, and status tracking. They make the closure more visible and better coordinated. They do not make it faster. They automate the management of the process, not the process itself.
ERP-native modules
SAP, Oracle, and their equivalents offer close management features, but these are typically bolt-ons to systems built for transaction processing, not close orchestration. They're also locked to a single ERP ecosystem — a significant constraint for any organisation with a mixed stack or that's gone through M&A.
Legacy RPA platforms
BlackLine, Trintech, and similar platforms built their core on reconciliation automation and close task management. They've layered AI features on top in recent years, but the underlying architecture was designed for rules-based automation, not agentic reasoning. There's a difference between having an AI feature and being built on an AI-native architecture.
The question to ask any vendor in this category: is your AI embedded in the workflow, or is it a separate module your team needs to interact with separately?
There's a difference between a platform that has added AI features and a platform built from the ground up to deploy AI agents in the close workflow.
1. End-to-end workflow automation — not just task tracking
A close platform should do more than tell you what's left to do. It should actively do parts of it. The benchmark question: can the system identify an unmatched transaction, propose the correcting journal, present it for approval, and post it — in one unbroken automated flow?
If the answer involves your team copying information between screens, the automation is incomplete.
2. Reasoning over exceptions — not just flagging them
Every reconciliation tool flags exceptions. The question is what it does with them next. A platform that flags and waits still leaves the hard work to your team. A platform that performs root-cause analysis, identifies likely resolutions, and presents them with supporting context is doing something meaningfully different.
Ask to see how the system handles a reconciliation break. Watch what it does with the exception — not just that it found it.
3. Human approval architecture — built in, not bolted on
This is the governance question, and it should be asked with precision. Ask: at which specific steps does the AI act without human approval? The answer should be: none, for any action that affects the ledger.
Preparation work — analysis, drafting, matching — can and should be automated. Posting, locking, and any action that changes a financial record should require explicit human sign-off. If a vendor can't give you a clear answer to which is which, that's a red flag.
4. Audit trail and SOX control integrity
The audit trail question is often asked but rarely asked precisely enough. The right questions are: Is every AI-proposed action logged? Are rejections and revisions captured alongside approvals? Is the evidence of who reviewed what, and when, exportable for audit purposes?
A system that logs completions but not the review process that preceded them is creating a gap in your controls, not filling one.
5. ERP-agnostic integration
Vendor lock-in to a single ERP ecosystem is a significant constraint that becomes more expensive over time — especially for organisations that grow through acquisition or operate mixed technology stacks. A close platform should read from and write to your existing GL and ERP, not replace them.
The corollary: a platform that requires you to replace your ERP before you can use it is not a close automation platform. It's an ERP sales pitch.
|
No |
100% |
Full |
Use these in your evaluation conversations. The quality of the answers will tell you more than any demo.
Show me a reconciliation break being identified, analysed, and resolved — end to end. How many screens does my team touch?
What happens when the system encounters a transaction type it hasn't seen before? Walk me through the exception handling.
Which specific actions require human approval before they affect the ledger? Can you show me the approval workflow in the product?
What does your audit trail capture — and what doesn't it capture? Show me what an auditor would see.
How does your system handle multi-entity groups with FX exposure? Is consolidation automated or manual?
What ERPs and GLs do you integrate with today? How long does a typical integration take?
What is live in production today versus on the roadmap? Can you show me a customer reference using the live features?
That last question is important. The financial close software market has a significant problem with roadmap selling — presenting planned capabilities as if they were available today. A vendor who can't point you to a production deployment with reference customers for the specific capabilities you need should be treated with caution.
If you're evaluating platforms that claim agentic AI capabilities specifically, the bar should be higher. Here's what genuine agentic architecture looks like versus AI features layered onto a legacy platform.
Genuine agentic architecture:
Specialist agents with defined scopes of responsibility (reconciliation, journal entry, variance analysis, consolidation, validation)
A Master Orchestrator that sequences agent activity, routes exceptions, and learns from each close cycle
Agents that can plan, execute, assess results, and adapt — not just execute a fixed sequence
Human approval gates at every material step, built into the agent workflow — not added as an afterthought
AI features on a legacy platform:
AI as a separate module or assistant that sits outside the core workflow
Recommendations that require manual action to implement
AI-powered' flagging without automated preparation or resolution
No clear articulation of which agent is responsible for what, or how exceptions are escalated
Ask any vendor: describe your agent architecture. If they can't name the agents, their responsibilities, and the escalation logic — it's a feature, not an architecture.
To set a concrete benchmark: a production-ready agentic close platform in 2026 should be able to automate the preparation work across all major phases of the close — reconciliation, journal entry, variance analysis, consolidation, and validation — while maintaining human approval gates at every step that affects the ledger.
It should work alongside your existing ERP and GL without requiring replacement. It should produce a complete, timestamped audit trail that satisfies external auditors. And it should be deployable without a multi-year transformation programme — starting with the highest-impact tasks and expanding progressively.
If a vendor can demonstrate all of that in a live product — not a demo environment, not a future roadmap — you're looking at something worth taking seriously.
The last thing worth saying to any finance leader evaluating this space: the technology is not the barrier it used to be. The harder question is organisational.
Finance teams that are piloting agentic AI in their close right now are not just compressing timelines. They're building something more valuable: institutional knowledge in how to govern AI-assisted financial processes. They're developing the internal fluency — the questions to ask, the edge cases to watch, the controls to verify — that will compound in value as this technology matures.
The teams that wait for agentic AI to become mainstream before they act will find themselves a full close cycle behind peers who moved earlier. In a function where speed, accuracy, and credibility are everything, that gap matters.
The window to lead is narrower than it looks.
0
8 April 2026 9:45:00 AM
8 min
The difference between a chatbot, a copilot, and an agent isn't just technical. It changes everything about what AI can actually do for your close.
If you've spent any time around technology in the last two years, you've heard the word AI used to describe everything from a spell-checker to a tool that can apparently run your entire finance operation.
Most of it is noise. But underneath the noise, something genuinely significant is happening — and if you're a CFO, Financial Controller, or finance leader, it's worth understanding what it actually is, because the practical implications for your close process are real and available today.
Let's cut through the jargon.
Level 1: AI Assistants (Chatbots and Copilots)
This is the AI most finance teams have already encountered. ChatGPT. A copilot embedded in your ERP. A tool that summarises documents, answers questions, or drafts commentary when you ask it to.
These tools are genuinely useful. But they share one fundamental limitation: they are passive. They wait to be asked. They respond to prompts. They don't initiate. They don't act. They don't follow through.
Ask a copilot 'what's driving the variance in this account?' and it will give you a useful answer. But it won't then draft the correcting journal, present it for your approval, and post it. That still falls to you.
Level 2: Automation Tools (RPA)
RPA sits at the other end of the spectrum. It acts, but it doesn't think. It can execute a sequence of steps reliably and at scale — as long as the steps are fixed, the data format is consistent, and nothing unexpected happens.
The closet is full of things that don't fit that description. Exceptions. Judgement calls. Intercompany disputes. Transactions that don't match any rule you thought to write in advance.
RPA hits these and stops. Or worse, processes them incorrectly and keeps going.
Level 3: Agentic AI
An AI agent is something different from both. It's designed to pursue goals, not just answer questions or execute fixed sequences.
An agent can plan a sequence of steps to achieve an objective. It can execute those steps. It can assess the results. And it can adapt — trying a different approach if the first one doesn't work, escalating to a human when it encounters something outside its confidence threshold, and learning from each cycle it completes.
An AI agent doesn't just tell you there are 235 unallocated transactions. It identifies them, proposes the right cost centres, summarises them for your approval, and posts the journals once you say yes.
The distinction isn't just technical. It changes what's actually possible.
There's a reason agentic AI is finding its footing in finance operations before almost anywhere else in the enterprise. The close process has a set of characteristics that make it exceptionally well-suited to this kind of automation.
It's high-volume and high-stakes
The close involves hundreds or thousands of individual transactions, journal entries, and reconciliation items — most of which follow recognisable patterns. That's exactly the kind of environment where an agent can add the most value: doing the pattern-matching and preparation work at scale, and surfacing only the genuine exceptions for human review.
It's sequential and structured
The close has phases. Each phase has dependencies. This is a natural fit for a Master Orchestrator — a coordinating agent that sequences work across specialist sub-agents, tracks progress, and routes exceptions to the right place.
The cost of errors is high and visible
A journal entry posted incorrectly creates audit risk, restatement risk, and the kind of conversation with your auditors that nobody wants. The stakes justify the investment in AI that can validate its own work — running 200+ integrity checks before anything posts, automatically, every cycle.
Finance leaders often worry that AI in the close means AI making decisions. The agentic model actually resolves this directly: the agent prepares, the human approves. Every proposed journal entry, every cost centre assignment, every period lock is gated behind an explicit human decision. The agent is an expert preparer. Your controller is still the decision-maker.
This isn't a compromise — it's by design. And it means the AI can take on the preparation work at scale without removing accountability from the people who need to hold it.
|
15 |
200+ |
7
|
Here's a concrete example of how this plays out.
It's day 2 of the close. The Recon Agent has automatically matched your bank feeds, sub-ledger, and GL. It's found 47 unmatched items. For 39 of them, it has identified the most likely root cause, proposed correcting entries, and queued them for controller review. For the remaining 8, it has flagged them as genuine exceptions — with context — for human investigation.
At the same time, the Journal Entry Agent has reviewed your accrual schedules, identified 12 accruals that need to be posted based on transaction patterns and your accounting policy rules, and drafted each one — with supporting rationale — for approval.
The Flux Agent has already started building the P&L narrative for the period, pulling context from your data to explain what drove each material movement. By the time the numbers are locked, the commentary is nearly ready.
Your controller reviews the queue. Approves 50. Adjusts 3. Rejects 1 with a note. Every decision is logged. Nothing posts without sign-off.
This is not a future state. This is how Octane FastClose operates today.
Let's address the question that comes up in every conversation about AI in finance: who is accountable?
The answer, in an agentic model designed properly, is the same person who was always accountable: your controller, your CFO, your finance team.
The agent doesn't hold authority. It holds preparation responsibility. Every material action requires a human decision. Every decision is logged with a timestamp, a user, and the supporting context that was presented at the time. Segregation of duties is maintained. SOX controls are intact.
What changes is where your team's time goes. Instead of preparing the journals, they're reviewing and approving them. Instead of building the variance commentary, they're stress-testing it. Instead of chasing reconciliation exceptions, they're resolving the ones that actually need judgment.
The agent does the work. The controller makes the call. That's not a slogan — it's the architecture.
— Octane FastClose design principle
The most common misconception about implementing agentic AI in finance is that it requires a wholesale transformation — replacing your ERP, restructuring your team, or undertaking a multi-year programme before you see any results.
It doesn't. The right approach is the opposite: start with the highest-impact, most automatable tasks in your existing close cycle. Get confident in how the agent performs, how your team interacts with the approval workflow, and what the audit trail looks like. Then expand coverage progressively.
Finance functions that are doing this now are already compressing their close by days, not hours. And they're building institutional knowledge in governing AI-assisted processes that will compound in value as the technology matures.
In the final post in this series, we'll get practical: what to look for in a financial close AI platform, what questions to ask vendors, and how to evaluate whether a solution is genuinely production-ready or still a roadmap promise dressed up as a product.
Ready to see an agentic close in action?
0
23 December 2025 10:09:01 AM
13 min
Hello Octane clients, partners, and friends — and hello to our own Octane crew reading this too.
As we wrap up 2025, we wanted to share a year-in-review that’s equal parts thank you, celebration, and “wait… did we really do all that?” 😄
This year was big. Big learning, big delivery, big growth, big laughs, and a few late nights fuelled by coffee, deadlines, and the occasional “just one more slide” moment.
So here it is: the story of Octane 2025 — what we achieved, what we built, and why we’re genuinely pumped about what’s coming next.
Then we headed to Vietnam for IBM’s APAC ecosystem gathering and came home with another major win — the IBM Partner Plus Award for SaaS (APAC).
What we loved most about this one is that it recognised outcomes: real transformation work, real client results, and consistent execution.
And yes… Orlando too. At IBM TechXchange 2025, Octane was named Runner-Up in the IBM TechXchange Excellence Awards.
Runner-up might sound modest, but trust us: on a global stage with the best of the best, it’s a massive achievement. It’s one of those moments where you quietly sit back and go,
“Okay… we’re in the game globally.”
This wasn’t a fluffy “look at our product” session. It was a proper transformation story about how we modernised a complex IBM Planning Analytics (TM1) platform — including:
Modern architecture and platform uplift
Model refactoring and performance focus
A foundation that’s now AI-ready
Reducing technical debt and improving user experience
Making the planning platform faster, more scalable, and actually enjoyable to use (yes, that’s possible)
Presenting alongside a client is the best kind of proof. Anyone can claim value — it’s different when your client stands with you and says,
“Yep. They delivered.”
One Team, One Family: Our Bangalore all Octane Meetup (and Mini-Meets)
2025 wasn’t just about client delivery and awards — it was also about us investing in the thing that makes Octane work: our people.
We held an All-Octane Meetup in Bangalore, bringing team members together from different cities and countries. For a globally distributed team, this stuff matters. It turns “workmates on Teams” into “people you can actually laugh with in real life”.
The week was a mix of:
Strategy sessions
Training and capability building
Sharing wins and lessons from the trenches
And a healthy amount of good food, banter, and team energy
We also had smaller mini-meets in Sydney, Mumbai, Gurugram and Hyderabad throughout the year — quick catch-ups, planning sessions, hiring and training meetups — the kind of moments that keep a team connected even as we grow.
If 2025 had a theme in the market, it was this:
CFOs are done with hype. They want what works.
So we leaned in — showing finance leaders how to modernise planning, reduce manual work, and actually take advantage of AI without risking governance, security, or sanity.
Across Melbourne, Sydney, and Mumbai, we participated in and hosted CFO-focused events, roundtables, and conference sessions where we covered topics like:
Modern FP&A and scenario modelling
Planning Analytics as a modern platform (not a “legacy TM1 box”)
AI for finance that’s practical, safe, and measurable
Agentic AI / orchestration and “digital labour”
Improving month-end processes and board reporting
What we loved most: the conversations were real. No buzzword bingo. Just finance leaders asking,
“How do we do this properly?”
…and Octane being able to answer with real experience and real examples.
We onboarded new clients this year and kicked off major transformation programs — across multiple industries and multiple countries.
Some of the work our team has been delivering:
Modernising IBM Planning Analytics platforms (on-prem → cloud / SaaS)
Rebuilding planning models to reduce complexity and improve adoption
Improving performance and governance for enterprise TM1 estates
Automating finance workflows and reporting
Building AI-ready foundations so clients can scale innovation safely
The common thread across these programs was simple:
Less manual work, faster insight, better decisions.
And as always, we aim for outcomes clients can feel — not just “a system upgrade”, but a real change in how finance operates day to day.
2025 was also the year our AI practice really took shape.
We’ve been talking about AI for a while (like everyone), but this year we focused on turning AI into:
Usable workflows
Secure and governed solutions
Real productivity improvements
Practical agents that operate inside enterprise environments
We built capability, grew the team, and expanded how we deliver solutions across IBM’s AI stack — especially around orchestration and automation.
For clients, this means AI that supports finance teams without creating chaos. And for our own internal team, it means we’re building a long-term competitive advantage — not chasing trends.
Our Octane Blue TM1 support services also grew significantly in 2025.
Support isn’t glamorous… until you realise what it really means for clients:
Your planning platform stays stable
Your users get answers quickly
Your changes don’t break everything
Your month-end isn’t derailed by a mystery data issue
You don’t have to scramble for talent every time something needs fixing
Octane Blue has become a key part of how we help clients keep their platforms healthy and continuously improving — not just “keeping the lights on”, but actively modernising over time.
If 2025 proved anything, it’s this: great support is a growth strategy, not an afterthought.
Now for the section we love because it shows who we are outside of spreadsheets and cloud architecture 😄
In 2025, Octane also sponsored and supported:
We also sponsored the WRX Car Club Junior Series, supporting young drivers and helping build confidence, skills, and community. It’s a big deal seeing juniors step up and improve — and it aligns perfectly with how we think about growth (in sport and in careers).
And yes — we also sponsored a squash tournament this year. Different arena, same energy. We love supporting community sport because it’s about participation, discipline, health, and momentum.
These sponsorships aren’t just marketing. They’re personal. They reflect our culture:
teamwork, improvement, friendly competition, and backing good people doing good things.
If we zoom out, 2025 wasn’t about one award or one project. It was about momentum.
It showed that Octane is:
Building global credibility (India, Boston, Vietnam, Orlando)
Delivering real transformation outcomes (not just implementations)
Expanding capability into AI and automation
Growing managed services (Octane Blue) with real depth
Investing in culture, team connection, and community
In simple terms:
we’re growing up — but we’re keeping our personality.
Now the fun part: what’s next.
In 2026, the market will shift from “AI curiosity” to AI accountability. People won’t ask “what can it do?” — they’ll ask:
“What results did it deliver?”
Our focus for 2026 is clear:
More measurable outcomes: less talk, more value
More AI that’s safe and practical: governed, secure, enterprise-ready
More platform modernisation: moving legacy environments to modern architectures
More client enablement: helping teams adopt and sustain change
More investment in our people: training, growth paths, and more meetups
We’re entering 2026 with confidence, momentum, and a bigger mission:
help finance teams run faster, smarter, and with less friction.
To our clients and partners: thank you for trusting Octane in 2025. Whether it was a major transformation program, support partnership, or a new engagement — we truly appreciate the opportunity.
To our team: thank you for the energy, the resilience, the late nights, the laughs, and the care you bring to every client outcome.
2025 was a massive year.
2026 is going to be even bigger.
From all of us at Octane — have an awesome holiday season, stay safe, recharge properly… and we’ll see you in 2026 ready to hit the ground running. 🎉
– Amendra & the Octane Team
0
17 December 2025 3:30:00 PM
10 min
In the last few months, the tech world has been buzzing with one phrase — Agentic AI.
From Stanford’s AI labs to Fortune 500 boardrooms, we’re seeing a powerful shift in how artificial intelligence is perceived and applied.
Until now, AI has been largely reactive — it answered questions, summarised text, generated images or code.
But Agentic AI takes a bold step forward. It doesn’t just respond — it acts.
Think of Agentic AI as a system that can understand a goal, plan a route, and execute tasks autonomously — almost like a digital teammate that thinks, collaborates, and delivers outcomes without waiting for line-by-line instructions.
Instead of relying on static workflows or pre-defined scripts, Agentic AI agents:
Plan their own actions based on context
Collaborate with other agents or apps
Adapt to changes mid-execution
And most importantly, learn from outcomes to improve the next run
It’s automation that thinks.
Every organisation is flooded with tools, dashboards, and systems — yet, productivity often remains locked behind repetitive manual actions: approvals, report refreshes, email follow-ups, reconciliations, HR queries…
Agentic AI changes that equation.
It creates a layer of intelligent autonomy on top of existing systems.
For example —
In HR, an agent can draft and send onboarding emails, create system access tickets, and schedule training sessions — all triggered by one hire event.
In Finance, an agent can reconcile accounts, generate exception reports, and even summarise insights for leadership reviews.
In Operations, field-attendance updates can trigger automatic payroll adjustments or rescheduling through integrated workflows.
These aren’t isolated bots. They are collaborating digital agents, working just like teams do — only faster, consistent, and 24x7.
While several players are entering the “agentic” space, IBM’s Watsonx Orchestrate brings a distinct advantage — enterprise-grade orchestration backed by responsible AI governance.
Watsonx Orchestrate isn’t just another automation tool. It’s a platform that lets you build, deploy, and manage AI agents that can connect across:
Microsoft 365 (Outlook, Teams, SharePoint)
ERP & HRMS systems
Collaboration tools like Slack or Asana
and even custom APIs within your enterprise stack.
Each agent can perform discrete actions — such as “summarise this email thread,” “update this record,” or “generate monthly KPI report” — and Watsonx Orchestrate links them intelligently into full, end-to-end workflows.
And the best part? It’s human-in-the-loop by design — meaning you stay in control while AI does the heavy lifting.
Agentic AI, powered by WatsonX Orchestrate, is the bridge between automation and autonomy.
It helps organisations:
Scale efficiency without scaling headcount.
Break silos between systems and functions.
Reduce turnaround times for repetitive but high-frequency tasks.
Empower teams to focus on decisions, not execution.
This isn’t just about “doing faster.”
It’s about redefining how work gets done.
Having spent years working with enterprise analytics and performance management platforms, I see Agentic AI as the missing layer that connects insight to action.
Tools like IBM Planning Analytics and Cognos give us the numbers and narratives — but it’s Agentic AI that can act on them:
It’s where analytics meets autonomy — and that’s where the next wave of enterprise productivity will come from.
Agentic AI is not a trend. It’s a fundamental evolution in how technology collaborates with humans.
Enterprises that embrace this shift early — especially through platforms like IBM WatsonX Orchestrate — will move from automating tasks to orchestrating outcomes.
And that, I believe, is where the real competitive edge lies.
0
2 August 2021 8:17:48 AM
8 min
It’s an amazing feeling as we gear up to celebrate Octane’s 5th anniversary. What started as an idea and nurtured by me and my senior leaders is fast transforming into an established and growing enterprise.
Five years is usually a significant milestone in a startup’s life as it pushes you into a grown-up zone. It’s been an exciting journey and the pandemic indeed threw some curveballs at us.
I am often asked about the motivation to resign from Macquarie Group to start Octane.
Macquarie was an excellent learning ground to nurture one’s entrepreneurial fire. I found a gap in the market to provide the latest AI-based technology to the Office of Finance. Some existing suppliers were either too expensive that meant the projects did not stack up against a cost/benefit analysis. Others simply did not have the skills or expertise to deliver complex Financial Projects.
My vision was to establish Octane as a specialist provider of Finance technology at a price point that allows for a wider adaptation. This was to be achieved using a hybrid onshore/offshore delivery and support.
I think we are keeping true to the vision and adjust and adapt due to market demands.
I must admit that the pandemic has validated our business model as the whole world adapted to remote working and sought out to optimise how they operate and a rapidly volatile environment.
Octane timeline over the last five years.
The biggest shoutout goes to the team we have at Octane, who are absolute champions and masters of their domain.
We have always sought to foster and build up the skills of our people. While Covid has put a stop to our legendary quarterly team get-togethers in different locations, we have slowly started planning to work towards our next get-together in Goa when restrictions lift and we are all able to travel.
It was amazing how our team got together as the effects of the pandemic were starting to be felt around the world. Being in a situation where we had our staff deployed to Dubai and Fiji, it was a mammoth effort and patience required to manoeuvre through lockdowns, quarantine and travel arrangements to get everyone back home safely.
As always, our main concern was the safety and welfare of our staff. During the pandemic, we have to provide different levels of support to our staff and their family as some of them were affected by the virus which has brought us closer together as a tight-knit team.
Work from home office became the norm in 2020.
Over the last five years, we have had the opportunity to work with clients of sizes and shapes. It’s pleasing to note that the vast majority of them are with us for the long haul, and we have an ongoing relationship with them. We have always sought to ensure that my team and I go above and beyond to ensure client satisfaction.
It has become our DNA.
Our get together with the Sydney Team before the pandemic.
I am also appreciative of my family and their patience. Starting out and building something like Octane requires immense time, effort and energy. There have been almost monthly travels (pre-pandemic) and lots of late-night calls.
The silver lining in the last 18 months was that I was able to spend a lot more time at home, and we managed to get some long-standing projects off the ground. It also allowed me to compete and win the NSW WRX Motorkhana championship, spend more time with kids and helped in their driving competitions. We have also started building a race car for my daughter.
My eldest daughter, Risha at her debut race at 14 - Wakefield Park Raceway.
In the last 18 months, we had a lot of time looking inwards and reflecting and adjusting our priorities. We, as Octane, has partnered with the charity B1G1. This allows us to support many charitable projects. We already have workshopped many creative ways to integrate our clients and tribe of followers to be part of our giving philosophy - i.e. provide a specific charitable project linked to each training or webinar you attend or when a client signs up for our Managed Services.
We have partnered with B1G1 during the pandemic.
The last five years have been exciting and did not follow the script. But we have adapted and evolved. It would be impossible for me to predict what the next five years would look like. One certain thing is that we will be strengthening our base in our core area of providing the best and latest technology to Finance teams.
We allow them to spend more time analysing and less time producing reports. The other strong theme emerging has been the signing up of clients in new markets like the USA and Europe.
With Covid taking away the barriers to remote work and most enterprises realising the value of having market insights in a timely manner could provide a significant competitive advantage.
Our Octane team looking forward to serving you for the coming years.
.png)
print
Leave a comment