UNPROGGED

AI Transformation Partner

Your AI quote.
Are you paying for
what you can’t see?

We open it up and show you what’s inside.
Then you choose: have us build it, or learn to build it yourselves.

AI Quote Second Opinion See our services

30 min · Online · Free | Already have a quote from another vendor? Bring it.

Scroll

Why Us

We read quotes like a builder —
because we are one.

unprogged doesn’t just write proposals. We build, we quote and we run what we build — ourselves. That’s why we can take someone else’s quote apart.

Estimate

We never quote a “lump sum”

Our own quotes are itemized by hours and rates. Because we use the same yardstick, we can take apart another vendor’s “lump sum.”

Product

We build and run our own products

Beyond client work, we build, sell and operate a character-based conversational AI agent. We see real development costs every day.

In-house

We’ve made teams self-sufficient before

Ten engineers were building agents in-house within three months. We know from practice where the line between in-house and outsourced work sits.

Our Promise

That’s why we can make three promises.

Define by Numbers

Define the results in numbers — then build

KPI design comes before development. We agree on the numbers to improve before we start. AI without a measurable target is self-congratulation — so we assess before we build.

End to End

Strategy through implementation, under one roof

No slide-deck-only proposals, and no building blindly to spec. Results die in the gap between design and implementation — so the same lead carries the work end to end.

Handover

Handed over ready for you to run

Source code, design documents, and runbooks are delivered as your assets. We transfer operation and maintenance to your team — including training, where needed.

Outcome — From Our Engagements

−80%

Maintenance cost reduction

CASE 01 | E-commerce · Operations AI agent

−50%

Reduction in site managers’ workload

CASE 02 | Construction-industry AI agent

3months

AI development brought in-house

CASE 03 | 10 engineers trained

01

The Problem

Can you put a number on your AI transformation?

Nobody can see what’s in the quote — or what the results will be — yet the project goes ahead anyway. AI projects that quietly fade away all look like this.

Symptom 1

Stuck at PoC

The demo worked. But no one can say whose hours will go down, or by how much. Six months later, no one brings it up.

Symptom 2

No one can explain the impact

The team says it “feels more convenient.” The board asks, “So what did it actually save?” Investments without an answer vanish in the next budget.

Symptom 3

Nobody uses it

The moment rollout becomes the goal, teams slip back into old habits. The only thing that changed is the license fee — a familiar story.

One root cause

The results were never defined in numbers.
So we show you everything — the numbers, and what’s inside.

02

Second Opinion

AI Quote Second Opinion

This is where to start. Bring us the AI quote you received from another vendor and we’ll read it the way a builder would. Give us 30 minutes before you sign.

01

Is the price fair?

We break it into hours and rates and check it against what building actually takes.

02

Where is it padded?

We open “lump sum” lines down to individual tasks and show you what is in them.

03

What could you do yourselves?

We separate what you can do yourselves from what is worth outsourcing.

STEP 1

Send the form

Choose “AI Quote Second Opinion” in the form below and send your name and contact details.

STEP 2

Share the quote

Once a time is set, send us the quote. Any format is fine — you may redact the vendor name or parts of the amounts.

STEP 3

A 30-minute call

From a builder’s standpoint, we tell you whether it is reasonable, where it is padded, and what you can do in-house.

  • 30 min · Free
  • NDA available
  • Vendor names can be redacted
Request a second opinion
03

Our Method

Measurable AI transformation, in five steps

If you decide to build, here’s exactly how we work: five steps, each with a defined deliverable.

STEP 1

Assessment

Break operations down to where, who, and how many minutes

Key activities

  • Interviews and on-site observation
  • Breaking down operations and measuring baseline hours and costs
  • Identifying which tasks to automate with AI

Deliverables

  • Process breakdown and baseline (Before)
  • List of tasks to automate with AI

Your decision

Choose the tasks to move into KPI design

STEP 2

KPI design

Agree first on what to improve and by how much. If we cannot set a number, we do not build.

Key activities

  • Agreeing on targets and success criteria
  • Defining how impact is calculated (e.g. hours saved × labor cost)
  • Scoping the PoC: range, timeline, and cost

Deliverables

  • KPI definitions and success criteria
  • PoC plan (scope and cost)

Your decision

Proceed to PoC, or stop

STEP 3

PoC validation

Build something that works within weeks, and measure it with real data

Key activities

  • Prototype focused on the target tasks
  • Validation with real data and real users
  • Before / After measurement
  • Surfacing accuracy, cost, and operational issues

Deliverables

  • Working PoC
  • Validation report (impact, cost, risk)
  • Design direction and estimate for the full build

Your decision

Proceed to full build, or stop

STEP 4

Build and rollout

Build the screens, workflows, and training your teams will actually use

Key activities

  • Production design and implementation
  • Integration with existing systems and data
  • Access control, logging, and security design
  • Team training and phased cutover

Deliverables

  • Production system
  • Runbooks and admin guides
  • Training materials and FAQ

Your decision

Go live

STEP 5

Measure and improve

Measure and report Before / After. Expand from what worked.

Key activities

  • Before / After measurement and reporting
  • Monitoring and incident response
  • Keeping up with model and spec changes
  • Rolling out what worked to other tasks

Deliverables

  • Impact report (hours, cost, KPI attainment)
  • Improvement and rollout proposals

Your decision

Continue or expand

Not “build, then measure.” Build to measure.

04

Security & Data

Security and data handling

The first questions anyone asks about AI adoption are built into our design from the start.

01

Your data is never used for training

We use configurations that keep your data out of external model training (API usage, training opt-out) as standard.

02

Clear data locations

Built in your own cloud (AWS / Azure / GCP) or within your environment. Data storage locations and flows are documented at the design stage.

03

Access, logs, and audits

Per-user permissions, operation logs, and audit reporting are standard items in every production build.

04

Confidentiality and ownership

NDAs are signed before we start. Project information and data are never used for other purposes, and deliverables and source code are handed over as your assets.

05

AI-specific risk controls

Designs include input validation, cited sources, and human review to guard against prompt injection, wrong answers, and leaks of confidential data.

06

Maintained in operation

Keeping up with model and library updates, fixing vulnerabilities, and reviewing access rights are part of ongoing maintenance.

Compliance with your internal security standards, such as ISMS, is confirmed during assessment and reflected in the design.

05

Service

Build it — or learn to build it.

Once you see what is inside, you will know what to outsource and what to bring in-house.

Core Service — Build

AI Agent /
AI System Development

Not build-and-forget contracting — development tied to KPIs. From requirements through implementation and operations, everything is done in-house, with accountability for the numbers. From full-scratch builds to integration with existing systems.

Beyond client work, we build and run our own character-based conversational AI agent platform. A company that builds is the one designing your AI transformation.

Core Service — Learn to build

In-House AI Program

We get your own team to the point where they can transform operations with AI. Hands-on support using your real work, not classroom lectures. We transfer development workflows and security standards too — so you are no longer locked into paying for every change.

Engineers or not — it doesn’t matter. Companies with a development team go as far as building AI agents in-house; companies without one get to the point where business teams can reshape their own work with AI tools.

Track record: 10 engineers brought in-house in 3 months, and many more

AI Transformation Consulting

AI Transformation Consulting

We start not from “what can AI do” but from “which number, in which process, do we change.” Assessment, KPI design, and PoC.

AI Technical Advisory

AI Technical Advisory

We assess, improve, and measure your AI development and usage from your side of the table — acting as your in-house AI lead, instead of leaving it to vendors.

06

Works

Case Studies

Across industries and company sizes, we didn’t stop at “installing” AI — we stayed until the numbers moved. Every result below is a Before/After measurement of a metric we agreed on before starting.

Case 01 | E-commerce company

Manual operations, handed to an agent

E-commerce operations, including web analytics, were manual and complicated. They depended on specific staff, and the maintenance burden kept growing. We broke the operations down, designed KPIs, and implemented automated execution and visualized results.

BEFORE Staff manually handle aggregation, reporting, and maintenance
AFTER Agents execute and visualize; people focus on review and decisions

Outcome

−80%

Maintenance cost

Case 02 | New construction-industry venture

A construction-industry agent, built from zero as a new business

Drawings, specifications, legal documents, construction plans, schedules — a document-heavy workflow was eating into site managers’ time across the industry. We designed and developed a construction-specific AI agent from the planning stage, through to rollout consulting.

BEFORE Checking drawings, specs, legal, and schedules consumes managers’ time
AFTER The agent handles first-pass processing; managers focus on judgment

Outcome

−50%

Site managers’ workload

Case 03 | Solution development company

From “using chat” to “building in-house”

AI use was limited to individuals chatting with AI. There was no integration into development, and no internal development workflow or security standards. We designed and ran a three-month hands-on program, from building the workflow to securing it.

BEFORE Individuals just asking an AI chatbot questions
AFTER The team builds agents in-house, with workflow and safety standards in place

Outcome

3months

10 engineers building in-house

Case 04 | VTuber operations

A fan-facing character that turns conversations into sales

We placed a fan-facing auto-responding character on a VTuber’s members-only e-commerce site. It introduces products in the flow of conversation and guides fans to purchase.

BEFORE The members’ site only listed products; fan contact was limited to streams and social media
AFTER The character responds on the site, and conversations lead to purchases

Outcome

+25%

Merchandise sales

Case 05 | Restaurant

Social media on autopilot — across multiple accounts

An owner-run restaurant had no capacity left for social media. We automated everything from creating posts to publishing, cutting the workload dramatically and letting the restaurant keep multiple accounts active on its own.

BEFORE One owner; social media updated only when there was spare time
AFTER Multiple accounts kept active with minimal effort

Outcome

+18%

Customer visits

Case 06 | Manufacturing

A second opinion on the incumbent vendor’s quote

The company was about to proceed with an incumbent vendor’s quote for new AI development. Using our second opinion before ordering, we broke down the “lump sum” lines and separated what could be done in-house from what really needed building — cutting the upfront cost substantially.

BEFORE A single quote from the incumbent vendor, with no way to verify the basis
AFTER Only the necessary development ordered; the in-house part done in-house

Outcome

−60%

Implementation cost

07

Our Products

Our Products

Alongside client work, we build and run products of our own. That’s where our feel for real development costs comes from.

Product 01

AI that runs your social media

SNS Contents Manager

Supports Instagram and X. It doesn’t just generate posts from a prompt: you set a persona (personality and role) for each account, and it generates posts informed by your knowledge base and past history.

  • A persona per account (personality, role, tone)
  • Posts informed by your knowledge base and past posts
  • Posts, views, follower growth and cost at a glance
SNS Contents Manager dashboard

Dashboard (sample data)

Product 02

A conversational AI agent, brought to life as a character

Character Conversational Agent Platform

An LLM embedded in a character that converses by text and voice and reacts with facial expressions and motion. Your own character (2D/3D) can be applied as is.

  • LLM control matched to the character’s personality, tone and role (two-way text/voice conversation)
  • Expressions and motion driven by detected emotion and intent
  • Answers grounded in your knowledge base — teaching materials, FAQs, manuals

Product example: teacher agent in an English-conversation product for children (released, in operation)

08

Company

Company

Company Profile

Company Name
unprogged, Inc.
Representative
Founded
April 19, 2022
Address
Garden City Tokorozawa 2F, 1-13-2 Kusunokidai, Tokorozawa, Saitama, Japan
Business
AI Agent / AI System Development
In-House AI Program
AI Transformation Consulting
AI Technical Advisory

A Message from Our CEO

The name “unprogged” is a coined term combining “unplugged” — meaning raw and authentic — with “program.”

Rather than offering cookie-cutter solutions, we tackle each client’s unique challenges with a hands-on, grounded approach — making sure real change takes root, even if it means doing things the hard way.

As AI continues to evolve, the demand for operational efficiency has never been greater.
Yet our mission is not simply to deliver AI-driven efficiency — it is to contribute to genuine, lasting improvements in enterprise value.

AI adoption is not the finish line.
We stay with you until the results show up in the numbers — then hand everything over so your team can run it. That is what AI transformation means to us.

CEO  

Contact

Contact

Consultations, estimates, or a free second opinion on another vendor’s quote — pick one and send us a note.

What can we help with? *