Data room · updated 6 August 2026

A card-sized AI-native device and service for the 2.7 billion people who work away from a desk.

Raising $200,000 for 5% · Post-money SAFE, $4M cap

Private. Please don't forward without asking — there are named counterparties on this page.

In one paragraph

Where this actually stands.

chit is a $129 machined-aluminium card with one button. Press it, talk, and chit turns what was said into records, expenses, calendar events and follow-ups — interpreted through who you are and what you do. Two SKUs are in production artwork with the factory and units have been produced. The store is live and takes real money through Stripe. Twenty co-branded devices are deployed with Flexbone, an AI healthcare company, with every capture resident in Flexbone's own VPC. We hold signed letters of intent from three more operators and roughly three hundred units of verbal interest, before we have spent a dollar on acquisition or opened preorders. We are raising $200,000 on a post-money SAFE at a $4M cap to fund the first production run and three revenue-share channel pilots.

~300
units of verbal interest: unsigned; 3
signed LOIs: non-binding; 20
devices deployed: Flexbone, live; 2
SKUs in production artwork: units produced

Traction

What is actually true today.

In the order a sceptic should read it: what is deployed, what is signed, what is only spoken, and what is built.

Deployed: Twenty co-branded chit devices in the field with Flexbone Solutions, each capture resident in Flexbone's VPC. This is a running deployment, not a pilot agreement.
Signed: Three executed letters of intent — Magnetism Motors (automotive retail), Stuart Pro HVAC (trades), and Flexbone (healthcare AI). All non-binding. Executed copies available on request.
Spoken: Roughly 300 units of verbal interest across the operators we have talked to. No purchase orders. No deposits. See the demand section below for the full caveat, which matters more than the number.
Built — hardware: chit one and chit one Pro in production artwork with the factory as of July 2026, drawn 1:1. Physical units exist and have been photographed for the store.
Built — software: Recording, transcription, context-framed summaries, the daily brief, and calendar and mail connections are running in the prototype app.
Commerce: hichit.com configures and buychit.com charges — a real Stripe Checkout session with the device as a one-time line and the membership as a subscription. End-to-end tested with a live card.
Revenue: Opentotal dollars collected to date, and from whom. If it is zero, this row says zero. A pre-seed round does not need revenue; it needs a founder who says the number.
Waitlist: Opensignups and preorders on hichit.com, pulled the week this link goes out.

Demand

Roughly 300 units, before we started.

~300

units of verbal interest from operators, ahead of launch, with no acquisition spend

Read this with the number

These are conversations, not orders. Nothing here is signed, no purchase orders have been issued, no deposits have been taken, and no delivery dates have been given. Every one of these operators can walk, and some will. We count them because they arrived without a marketing budget, a launch, or an open preorder page — from people who run the businesses chit is built for, several of whom have held the device.

Where the interest sits

Openthe per-organisation split of the ~300: who, what sector, how many units each, and how far the conversation has gone. Thirty operators at ten units each and one operator at three hundred are different companies with the same headline number, and a partner will ask which one this is.

What we would tell you not to believe

We have no conversion history to point at, so we are not going to hand you a percentage. The honest position is that unsigned verbal interest from a pre-launch hardware company converts at an unknown rate, and the first production run exists partly to find out what that rate is.

Letters of intent

Three signed. All non-binding.

Summaries below. Executed copies are available on request — we do not publish a counterparty's signed letter on a page they did not agree to appear on, and neither should anyone else.

Flexbone Solutions
Healthcare AI · flexbone.ai · Non-binding
Twenty co-branded chit devices, deployed and running, with all capture resident in Flexbone's own VPC and the intelligence customised to their operation.
Units
20 devices, deployed
Executed
Opendate executed
Signatory
Openname and title of the signatory
Commercials
Openis Flexbone paying today, and what for: devices, a per-seat fee, a pilot fee, or nothing yet? This is the first question a partner asks about the strongest logo on the page.
Magnetism Motors
Automotive retail · Non-binding
Openone sentence on what Magnetism Motors intends to do with chit: how many people, which part of the operation, and what problem they said it solves.
Units
Openunits or seats in the letter
Executed
Opendate executed
Signatory
Openname and title of the signatory
Commercials
Openprice, term, and whether a rev-share or a discount is attached
Stuart Pro HVAC
HVAC and trades · Non-binding
Openone sentence on what Stuart Pro intends to do with chit: crew size, which part of the job, and what problem they said it solves.
Units
Openunits or seats in the letter
Executed
Opendate executed
Signatory
Openname and title of the signatory
Commercials
Openprice, term, and whether a rev-share or a discount is attached

Openthe deck at invest.hichit.com lists a logistics LOI. It is not on this page. Either add it here with the same fields as the others, or correct the deck. And if the logistics counterparty is Triple C — the carrier one of the founders owns — it has to be labelled a related party right here, before someone finds it themselves.

A letter of intent is a statement of interest, not a contract. None of these obligate the counterparty to buy anything, and none should be read as booked revenue.

Case study

Flexbone runs on chit.

The clearest evidence that the enterprise motion is real: a company that took our hardware, put their own name on it, and pointed it at their own cloud.

Who they are: Flexbone Solutions (flexbone.ai) is an AI company working with healthcare customers. Their teams sit in rooms with those customers and the substance of the engagement — what was asked, what was promised, what was decided — lives in conversation.
The problem they had: Notes taken after the fact are the notes someone remembered, not the ones that were said. Across a team and a customer base, that is a compounding loss: nobody has a reliable picture of what is actually happening in the rooms, and the detail that mattered is gone by the time anyone needs it.
What is deployed: Twenty chit devices, branded as Flexbone's own hardware rather than ours, in the hands of their teams. One button, in the room, no laptop open and no app to launch mid-conversation.
Where the data lives: Every capture lands in Flexbone's own VPC. The audio, the transcripts and the briefs stay in their cloud — we do not hold them. The intelligence layer is customised to their operation rather than the consumer default, so the same sentence is read the way Flexbone's business reads it.
What it does for them: Their teams track notes across healthcare customers and pull data insights out of what was actually said in the room, rather than out of what someone typed up afterwards.

20
branded devices in the field: Their VPC
where every capture lives: Custom
intelligence, tuned to their work

What this does not yet prove

Opencontract value. Is Flexbone paying, how much, and on what terms? A twenty-device deployment at zero dollars is a strong product signal and a weak commercial one, and the difference is the whole question.
Openhow long it has been running, and daily active use across the twenty devices. Deployed and used are different words.
Openwhether Flexbone has said anything on the record we are allowed to quote, and whether they will take a reference call. A reference call is worth more than any paragraph on this page.
One deployment is one deployment. It shows the VPC architecture works and that a company will put its own brand on our hardware. It does not yet show that this repeats.

Why we lead with it

The enterprise version of chit is not a bigger consumer sale. It is co-branded hardware, the customer's own cloud, and an intelligence layer shaped to their business — a deployment a competitor cannot answer by shipping a cheaper recorder. Flexbone is the first instance of that motion, running, with real teams using it.

Product

What is built, and what is not.

Hardware

chit one: 53.9 × 86 × 3.9 mm, machined aluminium, ID-1 card footprint. $129.
chit one Pro: 57.9 × 88.1 × 6.7 mm, diamond-cut faceted face. $179.
Finish: One: bead-blasted natural anodised aluminium. Confirmed against photographs of the physical unit.
Factory: Openwho manufactures it, where, and what the relationship is: a signed agreement, a PO, or a quote. Name the partner or say you are not naming them and why.
Tooling: Openis the tooling paid for and cut, or quoted? What did it cost and who owns it?
MOQ and lead time: Openminimum order quantity and factory lead time from PO to units on a boat.
Certification: OpenFCC, CE, RoHS and Bluetooth SIG status and expected dates. An uncertified radio cannot legally ship in the US, and this is the item that most often moves a hardware launch by a quarter.

Software

Recording and transcription (Deepgram today)
Context-framed summaries — set at first open, per person and per trade
The daily brief: the day gathered into one page
Calendar connections — Google, Zoom, Teams
Mail connections and drafted follow-ups
Search across everything captured

Model-agnostic by design: transcription runs on Deepgram today and the architecture is BYOM, so an enterprise can bring its own model. The enterprise deployment puts capture, transcripts and briefs in the customer's own VPC.

Openwhich of the list above is running in the prototype for a real user today, and which is demoable but not in anyone's hands. The site states all of it in the present tense, so any gap between the two is a claim that needs correcting on hichit.com as well as here.

Recording, consent and privacy

chit has no wake word and no always-on microphone — the firmware cannot record without the button being pressed. That is a deliberate design decision and it is also the single biggest regulatory difference between chit and an always-listening wearable.

Openthe position on all-party-consent states. Roughly a dozen US states require every party to a conversation to consent before it is recorded. What does the product do about that — an in-app disclosure, a spoken prompt, a per-state setting, terms that put it on the user? A partner who has been near a recording product will ask this in the first call, and 'we have not looked at it yet' is a survivable answer. Not having one is not.Openretention and deletion policy, subprocessor list (Deepgram and whoever else touches audio), and whether the consumer product's data is used for training. hichit.com/privacy is live; this should agree with it.

Unit economics

The device is profitable on day one.

chit is not a razor-and-blades business. The hardware carries a real margin before the membership is counted, so the $9 is upside rather than recovery of a subsidised device. That is a materially different risk profile from a hardware company that has to keep a customer eighteen months to break even.

Per device

chit one — retail: $129
chit one — landed cost: $65factory price, freight, duty and packaging, per unit
chit one — gross margin: $64 · 49.6%our arithmetic on the two figures above
chit one Pro — retail: $179
chit one Pro — landed cost: Openthe $65 figure was given for a unit without splitting the two SKUs. The Pro is a larger body with a machined faceted face and will not cost the same.

Per membership

Price: $9 per month · $108 per year
Attach assumption: ~50%Plaud's own reported benchmark, not our forecast
Cost to serve: Openwhat one active member costs per month in transcription, inference and storage. Deepgram bills per minute, so this is knowable to the cent from the prototype's own usage. Without it the membership margin on this page is a guess.

First year, per buyer, at the ~50% attach benchmark

$129 device + (0.5 × $108 membership) ≈ $183 revenue

$64 hardware margin + (0.5 × $108 less cost to serve) — the second term is the NEEDS above

One customer, one year, at Plaud's reported attach rate. The multiplication is ours; the 50% is theirs. If chit's real attach lands at 25% the device still clears $64, which is the point of the paragraph above.

Acquisition

There is no CAC to report. Nothing has been spent on acquisition — the ~300 units of verbal interest and the Flexbone deployment came from founder conversations. The go-to-market plan is revenue-share channel partners rather than paid media precisely because DTC hardware CAC is where this category's margins go to die, and a channel partner costs nothing until they sell.

Sources: TechCrunch · Jun 2026 — Plaud ~50% attach

Use of funds

$200,000. Six months.

Go to market: $100,00050%; Channel partner pilots, vertical onboarding, the per-industry use-case work that makes each channel win compound.
Production run: $40,00020%; First run of 100 units, plus tooling, certification and packaging.
Buffer: $40,00020%; Unallocated. Hardware timelines slip and this is the line that absorbs it.
Software: $20,00010%; Infrastructure, transcription and inference cost, and the enterprise VPC deployment work.

Six months at the founders' current burn.

Openthe monthly burn number the six months is calculated from, and what is in it. $200,000 over six months is roughly $33,000 a month; if the real burn is $8,000 then the runway is two years and the ask should say so.Open100 units at $65 landed is $6,500 of product against a $40,000 production line. The other $33,500 is presumably tooling, certification and packaging — say which, with numbers. An unexplained 6× is the kind of thing that makes a careful reader re-check everything else on the page.Openthe first run is 100 units against ~300 units of verbal interest. That is a deliberate-looking decision and it should be stated as one: are you under-producing on purpose to validate before tying the round's cash up in inventory, or is 100 what the factory's MOQ and the budget allow? Both are fine answers. Not answering invites the worse assumption.

What the money buys

First 100 units shipped
3 revenue-share channel partners live
Attach rate measured against Plaud's reported ~50% benchmark
First membership ARR on the books

The plan

From tooling to trades in twelve months.

Q3 2026: Production run #1 · hero film · round closes
Q4 2026: Launch: DTC through hichit.com, first three revenue-share channel pilots live
H1 2027: Enterprise pilots — per-company apps and VPC deployments; attach measured against the ~50% benchmark
H2 2027: Scale the verticals that worked; seed round or profitability

The round

Post-money SAFE, $4M cap.

Raising: $200,000
For: 5%
Instrument: Post-money SAFE, $4M cap
Valuation: $4M post-money
Discount / MFN: Openstandard YC post-money SAFE with no discount, or is there one? MFN?
Pro rata: Openis a pro-rata side letter on offer?
Committed so far: Openhow much of the $200,000 is spoken for, and by whom. A round that is 40% committed reads completely differently from one at zero, and this is the number that creates urgency if it exists.
Target close: Openthe date you intend to close.

Cap table

Openthe pre-money cap table, at pre-seed simplicity: each founder's holding and percentage, any option pool, any existing SAFEs or notes outstanding with their caps, and anything owed to an advisor, contractor or early helper in equity. If it is clean — two founders, no pool, no prior instruments — say exactly that, because clean is an asset and it needs to be legible.

Corporate

Entity: Openlegal name, state and form. Delaware C-corp is what a SAFE assumes; if it is an LLC, that is a conversion to do before anyone can wire.
Incorporated: Opendate.
Founder vesting: Openis there a vesting schedule on the founders' stock, and has 83(b) been filed? An investor will not fund two founders with fully vested stock and no schedule.
IP assignment: Openare the designs, firmware and code assigned to the company in writing? Both founders have employers; that makes this the single most important corporate item on this page.
IP filings: Openany provisional patents, design patents or trademark applications, with filing dates. If there are none, say none — pre-seed hardware often has none and pretending otherwise is worse.

Team

We sell AI to the Fortune 500 by day — and run blue-collar businesses ourselves.

Shafay Ahmed

Co-founder · GitHub (Microsoft) · Georgia Tech CS

Senior Solutions Engineer at GitHub, on enterprise strategic accounts across Copilot, Advanced Security and Agent HQ; previously forward-deployed at Aisera and Couchbase. Serial builder — Keelway (freight TMS), TestDrivePro (dealership SaaS with paying customers) — and operator of an independent car dealership, which is exactly the customer chit serves.

Ahmad Agad

Co-founder · Keelway · Georgia Tech CS

Co-founder of Keelway. Owner-operator of Triple C, a carrier running over a hundred trucks — inside trucking, one of chit's launch verticals, every day.

Founder–market fit is not a slide claim here. We are the ICP.

Openare both founders full-time on chit, and if not, what is the trigger and the date? One founder is a Senior Solutions Engineer at GitHub and the other runs a hundred-truck carrier. That is the source of the founder–market fit above and it is also the first thing a partner will probe. Say the plan plainly: who goes full-time, when, and what this round funds of that.Openthe first hires this round pays for, if any, or a plain statement that it does not fund headcount.Openany advisors, and whether they hold equity. If there are none, say none.

Market

The deskless majority.

The full market case is in the deck. In brief: roughly 2.7 billion people — about 80% of the global workforce — work away from a desk, and the entire meeting-transcription stack assumes a calendar, a desk and a laptop. Bottom-up on US trades and field service alone: 12.5 million workers × ($129 device + ~50% attach × $108/yr) ≈ $2.3B serviceable per year, with trucking, dealerships and 1099 owners on top of that number rather than inside it.

Price anchors: Sandbar Stream is $249–299 plus $10/mo, Plaud runs $159–189 plus $8.33–19.99/mo. chit undercuts on the device at $129 and simplifies the service to one $9 plan.

The full market build-up, with sources, is slides 8 and 9 of the deck. invest.hichit.com

Sources: Emergence Capital — deskless workforce · BLS OEWS · May 2025 · TechCrunch · Jun 2026 — Plaud attach

Risks

What could go wrong.

Every one of these will come up in diligence. We would rather they came up here.

01
The verbal interest does not convert
Roughly 300 units of unsigned interest is the strongest number on this page and the softest. Nothing is committed, and pre-launch hardware interest converts at a rate nobody — including us — can currently quote. The first production run is sized at 100 partly to find out.
02
Certification and the shipping date
A Bluetooth device with a radio needs FCC and Bluetooth SIG certification before it can legally ship in the US, and certification is the item that most often moves a hardware launch by a quarter. Status is a NEEDS above, which is itself the honest signal.
03
Manufacturing concentration and tariffs
A single contract manufacturer and a US tariff regime that has moved repeatedly on China-manufactured electronics are both live exposures on a $65 landed cost. A tariff move eats hardware margin directly.
04
Recording law
Roughly a dozen US states require all-party consent to record a conversation. chit's press-to-record design is a better starting position than an always-on wearable, but it is a starting position, not an answer. See the product section.
05
The form factor is replicable
36Kr's own read on Plaud is that its form is easily replicable, and the same is true of ours. The defensible layer is everything the device feeds — compounding context, channel distribution, VPC enterprise deployment, BYOM — not the aluminium.
06
Founder time
Both founders currently run other things. That is the source of the founder–market fit and it is also a real execution risk until it is resolved on a date. See the team section.
07
A platform decides to build it
OpenAI has confirmed a family of devices and Amazon bought Bee. Our answer is that every funded player is building for the meeting-heavy desk worker, and the deskless majority has no incumbent — but a platform entering the vertical is a real tail risk and not one we can price.

Sources: 36Kr · May 2026 · The Next Web · Jul 2026 — OpenAI devices · TechCrunch · Jul 2025 — Amazon acquires Bee

Documents

Available on request.

Everything below exists and will be sent to anyone in an active conversation. Ask by email and it comes back the same day.

Executed letters of intent: Flexbone, Magnetism Motors, Stuart Pro HVAC — full signed copies.
The deck →: Fifteen slides, live at invest.hichit.com — it updates after it is sent.
Manufacturing drawings: 1:1 factory drawings for both SKUs.
The SAFE: Openis the instrument drafted and ready to send, and is it the standard YC post-money SAFE?
Cap table: Openas a document, in whatever form it exists.
Certificate of incorporation and bylaws: Openconfirm these exist and can be sent.
Bank balance: Opena partner writing the first cheque will ask what is in the account today. Have the number ready even if it is not on this page.

Contact

Ask us anything on this page.

Every NEEDS above is a question we will answer directly. Thirty minutes, whenever suits — the calendar is live.

Book thirty minutes ahmad@hichit.com

Fine print

1This page is confidential and is shared for the purpose of evaluating an investment in chit. Please do not forward or republish it — it names counterparties who agreed to work with us, not to appear on the internet.
2Nothing here is an offer to sell or a solicitation of an offer to buy securities. Any investment would be made solely through the executed transaction documents.
3Letters of intent summarised on this page are non-binding statements of interest. They are not contracts, purchase orders or booked revenue.
4Figures described as verbal interest are unsigned conversations. No purchase orders have been issued and no deposits have been taken.
5Forward-looking statements — the plan, the timeline, the unit economics and anything described as expected or projected — are estimates based on what we know today. They are not guarantees, and a pre-seed hardware company's estimates move.
6Product dimensions and finishes come from manufacturing drawings and finish samples and are pending final confirmation against the shipping unit.
7Some product imagery is a render or an AI-generated composition based on those drawings, shown while production photography is completed.

chit · ahmad@hichit.com · hichit.com · 15 sections · updated 6 August 2026

Where this actually stands.

What is actually true today.

Roughly 300 units, before we started.

Three signed. All non-binding.

Flexbone Solutions

Magnetism Motors

Stuart Pro HVAC

Flexbone runs on chit.

What is built, and what is not.

The device is profitable on day one.

$200,000. Six months.

From tooling to trades in twelve months.

Post-money SAFE, $4M cap.

We sell AI to the Fortune 500 by day — and run blue-collar businesses ourselves.

Shafay Ahmed

Ahmad Agad

The deskless majority.

What could go wrong.

The verbal interest does not convert

Certification and the shipping date

Manufacturing concentration and tariffs

Recording law

The form factor is replicable

Founder time

A platform decides to build it

Available on request.

Ask us anything on this page.