// MAKSINET LTD.— EST. 2007intake: 2 clients / quarter

AI transformation is a project not.
It's an operating model shift.

We don't sell consulting hours.
We become partners in the change itself.

  • ▸strategy → architecture → production, end-to-end
  • ▸outcome-based fees, owner-level accountability
  • ▸KPI accountability, P&L-measurable results
Start engagementcat ./manifesto.md
maksinet@partnership ~ %
> maksinet inspect --engagement standard

  model.........  outcome-based partnership
  fee...........  base + success share (KPI-tied)
  commitment....  12–24 months
  scope.........  strategy • architecture • production • scale
  status........  shipping ✓

> maksinet diff vs traditional-consulting

- hourly rates, slideware
+ shipped systems, P&L-measured outcomes
- hypotheses, POCs, "strategic recommendations"
+ live systems, KPI dashboard, accountability

> _

// our working stack

Anthropic
OpenAI
Google Cloud
aws
Vertex AI
Cloudflare
LangChain
Pinecone
Supabase
Vercel
NVIDIA
stripe
Deepgram
ElevenLabs
Anthropic
OpenAI
Google Cloud
aws
Vertex AI
Cloudflare
LangChain
Pinecone
Supabase
Vercel
NVIDIA
stripe
Deepgram
ElevenLabs
// 01

Manifesto.

RULE 01●

We don't sell hours.

We're outcome partners. Our fees are tied to your KPIs.

RULE 02●

We don't build POCs.

Day one we design production systems. Not demos — running software.

RULE 03●

We don't write decks.

No strategy memos. Architectural decisions and working code only.

RULE 04●

We don't run hour-long meetings.

Async, written, evidence-based communication. We respect your team's time.

RULE 05●

We don't just sell.

We think like owners, tell the truth, say 'no' when needed.

RULE 06●

We don't leave half-finished.

No engagement ends until scaling, training and handover are complete.

// 02

Capabilities.

ai-strategy()→

AI Strategy & Roadmap

Process inventory, ROI modeling, architectural vision. Where, when, in what order.

  • use-case mining
  • ROI model
  • build vs buy
  • tech selection
agents()→

Autonomous AI Agents

Production-grade agent systems. Voice, sales, internal ops, code generation.

  • voice agents
  • workflow automation
  • multi-agent
  • RAG
llm-infra()→

LLM Infrastructure

Model selection, fine-tuning, eval frameworks, production-grade LLM stack.

  • Anthropic
  • OpenAI
  • Vertex
  • self-hosted
data()→

Data & Operational Intelligence

Unify scattered data, real-time dashboards, predictive analytics. Shorten the decision loop.

  • data pipeline
  • real-time
  • vector store
  • analytics
cloud()→

Cloud & Edge Modernization

AI-native, scalable, FinOps-aware architecture on GCP, AWS, Cloudflare.

  • Vertex AI
  • Cloudflare Workers
  • Supabase
  • k8s
governance()→

AI Governance & Team

AI policy, security frameworks, internal training, Center of Excellence setup.

  • AI policy
  • compliance
  • training
  • CoE setup
// 03

Engagement model.

# No hourly billing. Three partnership models — all tied to outcomes.

/012 weeks

Discovery

Fixed fee

Process inventory, transformation map, first pilot scope. Required before any other engagement.

  • ▸Process map
  • ▸AI opportunity matrix
  • ▸ROI projection
  • ▸Pilot design doc
First step →
RECOMMENDED
/0212–24 months

Outcome Partnership

Base + success share

Our core engagement. Fixed monthly + KPI-tied success share. Outcome partnership. Limited capacity per client.

  • ▸Active engineering capacity
  • ▸KPI dashboard
  • ▸Monthly business review
  • ▸Handover & scaling included
Talk to us →
/03Long-term

Equity Partnership

Equity / revenue share

Founding partnership for AI-native company builds. Alongside the founding team — capital and engineering partner.

  • ▸Founding engineering
  • ▸Architectural leadership
  • ▸Equity stake
  • ▸Co-build model
Let's talk →
// 04

Our Products.

# Our own product portfolio — production-grade AI systems running at scale. We build with the same discipline for you.

850Santral
/850santralAI Phone Assistant

850Santral

24/7 AI phone assistant for businesses. Calls, appointments, reservations — marketing, sales, support, all powered by agents.

  • Voice AI
  • FreeSWITCH
  • AI Agents
Visit site
ClipBeat
/clipbeatAI Music Video Studio

ClipBeat

AI-powered platform that creates professional music videos in minutes. Studio, Agent, and Live modes with multi-track editor. 50K+ videos created.

  • AI Video
  • MiniMax
  • Fal AI
  • Kling
Visit site
Call2Me
/call2meVoice AI Platform

Call2Me

Voice AI platform. Build AI voice agents that handle millions of calls — build once, deploy everywhere. "Talk. Don't Type."

  • Voice AI
  • LiveKit
  • FastAPI
Visit site
Ikigai
/ikigaiAI Personal Coaching

Ikigai

AI-powered personal coaching app. 24/7 access, instant motivation, personal analysis, and mood tracking. iOS + Android.

  • AI Coach
  • Wellness
  • Mobile
Visit site
FlyBrand
/flybrandAI Marketing Autopilot

FlyBrand

Drafts 48 kinds of marketing pieces — ad copy to whitepapers — on schedule. For D2C, SaaS, creators, coaches, and agencies.

  • AI Marketing
  • SaaS
  • Automation
Visit site
Nina
/ninaAutonomous Desktop AI

Nina

Autonomous desktop AI for Mac. Reads emails, organizes files, runs automations — tell once, done forever.

  • Desktop AI
  • Mac
  • Automation
Visit site
// 04

How we work.

A meeting is not a tool — it's a cost. We work async, written, evidence-first — and that produces better work for both sides.

▢async-first

Async by default.

90% of communication is written (Slack/Linear/Notion). Meetings only for decisions or creative breakthroughs — never longer than 30 minutes.

▷loom-not-meeting

Loom > Meeting.

A 5-minute screen recording is more informative than a 30-minute zoom. Leadership watches at 2× speed, only stops where needed.

⊟writing-as-thinking

Writing is the certificate of thought.

Decision documents are written. An undocumented decision is an unmade decision. Every decision is searchable in your archive.

◷weekly-update

Weekly update — 1 paragraph + metric.

Every Friday: what we did, what we learned, what's blocking. You read it in 5 minutes, you know the week.

◧demo-fridays

Live demo every two weeks.

30 minutes. Working system, real data. No slides. Q&A included.

○single-thread

Single point of contact.

One decision-maker on your side, one lead on ours. We don't work with committees.

// Our meeting budget

  • Kickoff1 hour, once
  • Bi-weekly demo30 min × engagement length
  • Monthly business review60 min × 12
  • Emergency decision call30 min, on demand
  • Everything elseasync, written
// 05

What the first 12 weeks look like.

Typical schedule for Discovery + Pilot. Outcome partnership is built on top of this.

  1. Week 0
    Mutual NDA + access

    NDA signature, repo/data/system access, initial documentation share. Async, done in 2 days.

  2. Week 1-2
    Discovery: deep dive

    Processes, data, team, current stack. 6-10 hours of interviews (users, leadership, eng). Output: transformation map.

  3. Week 3
    Pilot design

    Architectural decisions, KPI definitions, success criteria, contract finalization. SOW signed.

  4. Week 4-6
    Pilot build

    Production-grade, real data, real users (limited). First demo in Week 6.

  5. Week 7-9
    Pilot testing & iteration

    Live with limited users, KPI measurement, fast iteration. Go/no-go decision in Week 9.

  6. Week 10-12
    Scale or handover

    If pilot succeeds: roll out organization-wide. If not: handover & honest report.

// 06

We won't work with you if....

Filtering saves both your time and ours. There are better partners for the following profiles:

×

You want to 'have done' AI on paper.

Need a leadership presentation, no real transformation needed. Our model burns money in this scenario.

×

You require hourly billing.

Your procurement may force this. There are workarounds, but if you have no flexibility, we're not your fit.

×

You decide by committee.

A 9-person decision board = a 9-month project = zero momentum. A single decision-maker is required.

×

You don't want us touching your data.

AI transformation doesn't happen without data. We can't help if you only want a UI layer.

×

You expect ROI in the first 6 months.

Pilot is 8-10 weeks. Scale is 6-9 months. Sub-6-month ROI expectations aren't realistic.

×

You're chasing AI hype with budget.

If budget is allocated to tech, not to a problem, the engagement spirals. Problem first.

// 08

Case studies.

# Three anonymized cases, three industries. Specific metrics. One-paragraph summaries.

/telecomOutcome Partnership

A voice agent that resolves 63% of first-level calls without ever reaching a human operator

Grew per-agent monthly call capacity 2.3× while raising CSAT from 4.1/5 to 4.3/5.

63%
First-level resolution rate
2.3×
Calls per operator capacity
4.1 → 4.3
CSAT score
Read the full case
/e-commerceOutcome Partnership

A RAG assistant grounded in the live product catalog that resolves 47% of support tickets without human intervention

The same assistant produced a +18% lift in product search conversion — turning the engagement into 3× revenue uplift versus support cost savings.

47%
Tickets resolved without human handoff
+18%
Search → add-to-cart conversion
+9%
Overall basket conversion
Read the full case
/b2b saasEquity Partnership

A multi-role AI agent platform that unifies customer onboarding, support triage, and invoice reconciliation

Cut the ops team's weekly workload from 11 person-weeks to 3.2; savings reinvested into ARR growth.

11 → 3.2
Weekly ops person-weeks
$420K
Avoided hiring cost (12 mo)
92%
Onboarding 'first-week value' rate
Read the full case
See all cases →
// 07

The people behind it.

UÇ
Umut Çetinkaya
Founder · Engineer

18 years of product and engineering. Lead engineer and founder roles in shipping multiple products from zero to production. Last 3 years deep in generative AI systems — LLM agent platforms, voice AI, multi-agent orchestration. Positions Maksinet as an engineering firm, not a consultancy.

umutcetinkaya.com →

// Data & security principles

  • 01.NDA + DPA standard

    Mutual NDA before discovery. GDPR / KVKK-compliant DPA framework.

  • 02.Your data stays in your cloud

    Where possible: private deployment, on-prem inference, models running in your own VPC.

  • 03.Audit trail for everything

    Every LLM call, every decision, every access logged. Fully visible to the client.

  • 04.Code ownership is yours

    All code and architecture written during the engagement is contractually yours. No vendor lock-in.

  • 05.Exit strategy from day one

    Handover documentation begins in week one. You never end up dependent.

// 08

Outcomes.

# KPIs measured on the P&L, defined together with our clients.

4×
operational throughput
average client outcome
60%
manual workload reduction
in the first 6 months
3×
feature ship velocity
via AI-native flow
+35%
sales conversion uplift
in AI-assisted flow
8 wks
time to production
discovery included
0
POC graveyard
no work goes unshipped
// 09

FAQ.

01.Why don't you bill by the hour?+

Hourly billing incentivizes us to work slowly, and you to extend meetings. Outcome-based aligns us on the same goal: fast, correct, durable results.

02.How long do you typically work?+

Discovery is 2 weeks. Outcome partnership is 12–24 months. Equity engagements are long-term. We don't do one-off 'build a POC and leave'.

03.What kind of clients?+

Maximum 2 new clients per quarter. Usually: ready to scale revenue, decisive leadership, budget allocated for transformation. SMB to enterprise.

04.I have an internal team — how does this work?+

That's the ideal setup. We work alongside your team, transferring knowledge. By engagement end your team can build independently.

05.What technologies do you go deep on?+

Anthropic, OpenAI, Google Vertex AI, AWS Bedrock; agent frameworks; voice/vision pipelines; Cloudflare/edge compute; modern data stack. But: problem first, tech second.

06.How do you handle data and security?+

NDA and DPA standard. Sensitive data stays in your cloud (private deployments, on-prem inference, GDPR/KVKK-compliant pipelines).

~/start-engagement

// 10 — let's start

We'll share our first read in 30 free minutes.

We listen to your current processes and give you our first opinion on where to start your AI transformation. No commitment, no surprise invoice.

// Click any button below — your email client opens with a pre-filled draft. Just answer the questions and send. This way we don't waste time in the first call.
email
hello@maksinet.com
schedule a call
30-min discovery call

$ response_time: < 24h · $ nda: optional · $ initial_call: free