Sobin George Thomas
Founder & CEO, ArkOne AI
Sobin George Thomas is the founder and CEO of ArkOne AI. He advises enterprise organizations on AI transformation — the structural decisions around governance, operating model redesign, and capability building required to move beyond pilots to measurable, compounding outcomes. His work spans banking, insurance, consumer goods, and engineering clients, where ArkOne engagements have delivered results like a 67% reduction in analyst review time and 2,400 hours saved per quarter.
专业领域
67 篇文章
Real estate sales engine
How do you actually tell a buyer from a browser in three questions, so that routing them happens by itself?
One qualification call runs about five minutes, timed by the agent who publishes his script. Nobody has published how many enquiries are browsers.
Real estate sales engine
How do you actually build the pitch from what you already know, so that the negotiation starts ahead?
$33,000 a year is the US national average for the listing coordinator whose duty list names the comparative market analysis.
Real estate sales engine
How do you actually write fifty messages that sound like you on a Tuesday, so that sending them is one button?
Fifty messages at ten minutes each is a working day, so the template wins and the owner gets the same letter as the street.
Real estate sales engine
How do you actually message fifty owners on WhatsApp in a morning, so that your number still works on Friday?
Fifty owners by approved template, charged only on delivery, at the rate Meta prints. The cost with no line on the card is the number itself.
Real estate sales engine
How do you actually find the person behind the holding company, so that reaching them happens by itself?
Five products sell the human behind a holding company by quote only, and not one prints a price. A name from the company's own page costs nothing.
Real estate sales engine
How do you actually keep an agreed deal moving, so that nothing dies in silence between offer and signature?
A UK agency pays £27,000 to £30,000 a year for the person who chases agreed sales to completion. The US pays $39,000 for the same seat.
Real estate sales engine
How do you actually keep the CRM true without typing a word, so that the Monday number is worth reading?
No agent has published the hours the typing takes, so this page prints none. A contact with nothing written on it starts the next call blind.
Real estate sales engine
How do you actually keep the twenty worth a call and say why, so that finding the human happens by itself?
One recorded-document lookup per asset, then the same lookups next quarter because nobody wrote the verdict down. Score and reason go beside the row.
Real estate sales engine
How do you actually list every owner in your patch from the public record, so that sorting them is automatic?
Four hundred assets is four hundred registry searches before sorting can begin, and the name that comes back is often a company, not a person.
Real estate sales engine
How do you actually match what they said to what is on your books, so that telling them happens by itself?
The vendor's own match processor asks four decisions every run, after three clicks to reach it. Twenty sent and none liked has never been timed.
Real estate sales engine
How do you actually name the one problem people call you for, so that the list of who has it writes itself?
Nothing in the category does this stage: every tool starts at a filter form you have already filled. The sentence comes from your own past work.
Real estate sales engine
How do you actually make sure no enquiry dies unread in the team, so that the viewing books itself?
A US portal takes 30 per cent of a commission at closing for routing an enquiry, 35 per cent above a price line. That is this stage left unbuilt.
Real estate sales engine
How do you actually pick the channel they answer on, so that every reply lands in one inbox?
An owner who messages you opens a 24-hour window in which every reply is free. Outside it the same reply is a priced template.
Real estate sales engine
How do you actually reach them without a credit bureau, so that the first message is one you can defend?
$53,088 is the fine the FTC prints per violation for calling a US number on the registry. Every number gets its source and its basis beside it.
Real estate sales engine
How do you actually stop viewings where nobody turns up, so that the follow-up happens by itself?
One Florida broker bills the other side $250 for a viewing nobody attended. Her count, never a market rate.
Real estate sales engine
How do you actually stay in front of a client who buys in two years, so that they call you and not the portal?
The trade calls a lead back one and a half times after first contact, measured across 384 brokers. By email, twice.
Real estate sales engine
How do you actually answer the enquiry that lands at eleven at night, so that it is qualified by morning?
917 minutes was the average wait before a broker replied across 384 brokers in eleven states, and nearly half of the enquiries were never answered.
Real estate sales engine
How do you actually run the whole first conversation on WhatsApp, so that the viewing books itself?
Every reply inside WhatsApp's customer service window is free, and the customer resets the clock each time they answer. What it costs by hand is an evening.
Real estate sales engine
How do you actually know which document is missing and from whom, so that the file closes itself?
$39,000 a year is the US national average for a transaction coordinator, hired to know which document is missing. Without the hire, it is your evenings.
Real estate sales engine
How do you actually see which stage is leaking on Monday, so that Tuesday is different?
No hours figure for this stage was verified anywhere, so this page prints none. What is paid is Tuesday, decided from a number nobody joined.
GTM engine
How do you actually run a content calendar that publishes without you, so that capture happens by itself?
A scheduling tool buys you the timing and leaves you every other job. Four rules in front of the publisher make an unattended calendar safe to run.
GTM engine
How do you actually keep a pipeline that is true on Monday, so that the forecast writes itself?
Typing a stage pays the typist nothing, so it gets dropped first. Stop storing the stage, store the facts, and derive it each time the pipeline is read.
GTM engine
How do you actually run a weekly review that changes something, so that next week starts by itself?
Most reviews spend the hour on outputs nobody can act on. Twenty minutes, input numbers only, and one dated decision the tool will not let you skip.
GTM engine
How do you actually turn that sentence into an ICP a machine reads, so that channel choice happens by itself?
A tier column typed from memory decides nothing. Stored as an object with a negative gate, the buyer becomes a score every later stage can read.
GTM engine
How do you actually book meetings without the back-and-forth, so that the pipeline stays true by itself?
Scheduling is close to free. A booking that moves the stage, writes the brief and notices a no-show is the half nobody sells, and the half the pipeline reads.
GTM engine
How do you actually capture a reader into a system, not an inbox, so that the CRM fills itself?
The click knew where it came from. A form that emails you throws that away, and no later repair exists, because the browser that knew has closed.
GTM engine
How do you actually forecast from the pipeline without a spreadsheet, so that one dashboard fills itself?
The probability column is typed from memory, multiplied, and presented as data. A stored number with its reasons can be reviewed three weeks later.
GTM engine
How do you actually make the CRM fill itself, so that the sequences start by themselves?
A CRM that is full and untrue is worse than an empty one, because people act on it. One write path in, and a stage derived from facts, fixes both.
GTM engine
How do you actually pick the two channels that matter, so that the content calendar fills itself?
Five small experiments, each stopped before it could say anything. Counting leads by channel with a total beside each one turns the choice into arithmetic.
GTM engine
How do you actually run sequences that stop when a human replies, so that a meeting gets booked by itself?
A person answers, the sending tool marks the thread finished, and the rest of the company carries on as though nothing was said. The stop has to become a fact.
GTM engine
How do you actually write the one sentence that decides who you sell to, so that the ICP writes itself?
A market definition that lives in one head drifts between Monday and Thursday. Written as one sentence with an exclusion, it becomes a gate later work reads.
GTM engine
How do you actually see the whole engine on one screen, so that the weekly review runs by itself?
A missing number looks exactly like a low one. One event writer, one number per stage and a total beside each count is what makes a screen worth reading.
Outbound engine
How do you actually find the one person who can say yes, so that getting their email happens by itself?
A bought list hands you a title with no provenance. A name from the company's own leadership page held 12 times out of 13; an aggregator name, 0 of 17.
Outbound engine
How do you actually follow up without remembering to, so that logging it all happens by itself?
A lead does not carry dates. It carries its messages in order, and the next one is the first not yet sent. A reply stops it, out of office pauses it.
Outbound engine
How do you actually get their email without guessing, so that writing to them happens by itself?
A guessed address that reaches a stranger is a complaint against your domain. Every address carries its source, and accept-all is a verdict of its own.
Outbound engine
How do you actually hand a warm reply to a call, so that reading the weekly number happens by itself?
Booking is solved for $10 a month. The brief the caller needs is sold by nobody, so Claude Code writes it from the timeline the moment a booking lands.
Outbound engine
How do you actually keep only the companies that fit, so that finding the right person happens by itself?
A list cut without reasons is cut again next quarter. Scored by Claude Code with reasons in fields, every company is kept or archived once, in writing.
Outbound engine
How do you actually list every company in your market, so that filtering them happens by itself?
A list with no total beside it looks complete. Listed from free sources by Codex, every company arrives with its domain, its source and the count.
Outbound engine
How do you actually log everything without typing, so that the handoff to a call happens by itself?
The CRM is true for an hour after the Friday update. Codex writes every send, reply, stop and step to the lead as it happens, in the words it happened in.
Outbound engine
How do you actually read and sort every reply, so that the follow-ups happen by themselves?
An agency hires a person to read replies all day. Codex matches each reply to its send, gives it one of seven labels, and files it before you read it.
Outbound engine
How do you actually read the weekly number, so that fixing the leaking stage happens by itself?
A stage that means a feeling cannot be counted. One screen every Monday: twelve counts with totals, the conversion between each pair, and the stage that leaked.
Outbound engine
How do you actually send without getting banned, so that reading the replies happens by itself?
Bursts are what the platforms watch for. Email goes out under a per-mailbox budget that doubles weekly and halves after a bounce day; LinkedIn stays by hand.
Outbound engine
How do you actually write down your market, so that listing every company in it happens by itself?
A market that lives in the founder's head cannot be checked. Written as a file with edges, it becomes the search the next stage runs on its own.
Outbound engine
How do you actually write one message per person, so that sending on a schedule happens by itself?
By the thirtieth message the hook is the company tagline. One generator, one verifier and one sweep, and the run passes only when both counts read nought.
Aviation
A Route Model, Asked Blind: AI Route Selection for Regional Indian Aviation
We built a route-selection model for regional Indian aviation from one public dataset. Asked blind, it placed 8 of a carrier's 10 routes in its top 100 of 2,743 city pairs.
AI Agents
Govern What Your AI Agents Send: Two Filters, Not One
Producing text and code has become close to free; judging it has not. Firms that fuse integrity control with expressive judgement slow the first to protect the second, right as their own agents start sending more than their staff do.
AI Agents
Sales Follow-Up: The First Process an AI Agent Should Own
Reps sell less than 30% of the week; the rest leaks into follow-up, routing, and CRM hygiene. That leak is the most measurable first process an AI agent can own.
AI Agents
Which Work Can an AI Agent Own? Run the Verifiability Test
Dubai is backing agentic AI adoption across 295,000 companies. The firms that benefit will be the ones that pick the right work to delegate, and the sorting rule is not sensitivity. It is whether the outcome can be checked by something the agent does not control.
AI Agents
Why AI Agents Cheat, and the Governance That Actually Holds
52% of DIFC firms now use AI, and 21% of those using it in critical processes lack clear oversight of it. The gap matters because capable agents game their objectives, and you cannot audit one by asking it what it did.
AI Strategy
Data You Never Collected Cannot Be Subpoenaed
Databases built for one purpose get repurposed. That is the observed default, not a prediction. Data minimisation has been sold to executives as a compliance cost; it is more accurately a purchase of optionality across every future where the data becomes a liability.
AI Strategy
Every AI Judgement Your Firm Stores Is a Future Exhibit
A KYC score, a fraud flag, a CV rejection: each is a durable, timestamped judgement about a named person or company. The reasoning that produced it evaporates in months. The record lasts years, and one day someone with authority will ask who signed it.
AI Strategy
The AI Business Case That Gets Funded Destroys the Upside
Tight scope, named benefit, defensible measurement: every instinct that gets an AI investment approved also forecloses its largest return. The by-products your programme will depend on in three years are being priced at zero today.
AI Strategy
The AI Layer You Own Outlives the Model You Rent
A supply shock does not reveal scarcity. It reveals which of your capabilities were leased. Every AI roadmap now terminates, several layers down, in someone else's policy, and the firms that survive substitution are the ones that kept their calibration assets portable.
AI Strategy
The Automation Your Capex Committee Can Still Approve
Trade policy has moved industrial robots from capital equipment into politically volatile territory, and payback periods are dying in committee. Process automation runs the other lane: no tariff, no landed cost, and it ships on the AI subscriptions you already pay for.
AI Strategy
Why 95% of AI Pilots Never Touch the P&L
MIT's NANDA study found 95% of generative AI pilots produce no measurable business return. In the GCC, 63% of organisations are still pre-implementation. The constraint was never the model. It is where the work waits.
AI Strategy
Your AI Vendor's Org Chart Is the Roadmap You Were Not Shown
Every frontier-lab reorganisation of the past three years was triggered by a shipping failure, not a research one. The state of your vendor's org chart is a direct input into the roadmap you were sold, and you have almost certainly not priced it.
AI Strategy
Five Principles of AI-Transforming Companies
WEF research on 450+ executives reveals five principles shared by the 15% transforming with AI. Organizational discipline separates leaders from followers.
Workforce
Why Your Job Title Will Be Irrelevant
Unilever unlocked 500,000 hours of capacity without hiring anyone. Moderna merged HR and IT under a single executive. Yum China fills 89% of restaurant roles in 1–2 weeks. AI is turning talent management from role-based HR administration into a dynamic capability system.
AI Strategy
The End of the Annual Plan
Canada Goose cut planning cycle time by 60% and improved revenue forecast accuracy by 4%. S&P Global analysed 190,000 earnings call transcripts to extract signals no analyst had spotted. The annual planning cycle is being replaced, not accelerated.
Innovation
How AI Is Compressing the Decade-Long R&D Timeline
A drug that fails in Phase III trials has consumed $500M–$1B. AI identifies the same failure at the hypothesis stage, when total spend is under $100K. Nearly 40% of executives now rank R&D as the top function benefiting from AI investment.
Operations
The Factory of the Future Runs on Signals, Not Schedules
2.4× greater productivity. 40–60% lower energy consumption. 27% reduction in order lead time. The WEF's 2026 study shows AI-enabled operations running on a fundamentally different architecture, not an incrementally better one.
Enterprise AI
AI Is Redesigning Customer Experience From Scratch
Up to 25% higher conversion, 21% churn reduction, 5–8% revenue uplift. The WEF's study of 450+ executives shows these as the median outcomes from organizations that have moved CX from static journeys to real-time adaptive systems.
AI Strategy
Why Only 15% of Companies Are Transforming with AI
The WEF surveyed 450+ executives across industries. Only 15% have moved beyond pilots to fundamentally redesign how their businesses operate. Those 15% report 2.4× greater productivity and 2.5× higher revenue growth. The gap is compounding.
AI Strategy
Most AI Programmes Don't Fail. They Stall.
The majority of enterprise AI investments don't end in visible failure. They produce one good pilot, then nothing. Here is why, and what the organisations that break through are doing differently.
AI Strategy
Why the CFO Is Now the AI Decision-Maker
87% of CFOs say AI will be critical to their operations in 2026. Yet only 12% report both cost and revenue gains. The gap is in who steers the spend, and how, not in the technology.
AI Strategy
The Integration Tax: What Enterprise AI Really Costs
The model is not the expensive part. For most enterprise AI programmes, integration into legacy infrastructure accounts for 40–60% of total delivery cost — and most business cases never count it.
AI Strategy
AI Readiness Is Not a Technology Question
Most organisations approaching enterprise AI frame readiness as a question of infrastructure and tooling. The evidence suggests it is primarily a question of process clarity, data governance, and decision architecture.