01 · Before
SaaS turned your process into settings.
Anything the menus can’t express, you live without.
Build the logic. Buy the platform.
Build vs. buy, in the age of AI. AI makes business logic cheap to write. Running the platform underneath it is still a full-time job. Own the first. Buy the second.
Based on the position paper by Binny Gill, Founder & CEO
built by your builders
SaaS
Business logicYour business logicthe vendor’s versionEnglish as Code · yours
Security · Governance · Audittrust
Scaling · Reliability · SLAsrun
Deployment · Cloud · Kubernetesrun
Logging · Metrics · Monitoringsee
the platform · Kognitos
The shift
AI splits the SaaS bundle in two
01SaaS sold you a platform you couldn’t build, wrapped around logic you couldn’t control. Now you can own one and buy the other.
02 · Now
AI lets you write it in plain English.
Working logic in hours, readable by the business.
03 · Still true
Running a platform didn’t get cheaper.
Agents write code. They don’t carry the pager.
The platform
Build what makes you different. Buy everything underneath.
02Your builders write the rules. Kognitos carries the 14 guarantees those rules need in production, so nobody on your team has to run a platform.
You build
Your logic
The rules that make your business different, written in plain English by your team.
Examples
- Invoice exceptionsFreight under 2% from a trusted vendor? Accept it. Otherwise, ask the buyer why.
- Vendor onboarding
- Cash application
- Claims intake
- Any rule you can describe
You buy
Kognitos
Everything underneath your logic, run for you in production.
Included
- Write itAuthoring, versioning and the logic runtime
- Do the workConnectors, document understanding, people for exceptions, model management and evals
- Trust itSecurity, governance and permissions, audit and replay, compliance
- See itLogging, metrics and monitoring
- Run itDeployment, scaling, reliability and SLAs, orchestration and state
Build it yourself and you also sign up for:
- 24/7On-call rotationFor as long as the logic runs.
- 1×/yrCompliance auditControl testing and remediation, every year.
- 14Guarantees to staffEvery one of them, if you build the platform.
See what each guarantee costs you every year
| Guarantee | What you own, every year |
|---|---|
| Run it | |
| Deployment and environmentsCloud accounts, Kubernetes, CI/CD, environment promotion, rollback | Upgrades, cluster maintenance, release engineering |
| Scaling and performanceAutoscaling, queues, back-pressure, load and soak testing | Capacity planning, performance regressions |
| Reliability and SLAsRetries, failover, idempotency, incident response, status reporting | 24/7 on-call, incident reviews, SLA reporting |
| Orchestration and stateLong-running runs, scheduling, pause and resume, durable state | Stuck-run triage, migrations of in-flight state |
| Trust it | |
| SecurityAuthentication, authorization, secrets, encryption, tenant isolation | Patching, penetration tests, vulnerability response |
| Governance and permissionsRoles, approvals, change control, applicability scopes | Access reviews, policy changes |
| Audit and replayDecision records, versioned rules, replay of past runs | Evidence requests, retention, auditor walkthroughs |
| ComplianceSOC 2, HIPAA, ISO 27001 controls and evidence | Annual audits, control testing, remediation |
| See it | |
| Logging, metrics, and monitoringStructured logs, metrics, tracing, alerting, dashboards | Alert tuning, observability costs |
| Do the work | |
| ConnectorsERPs, email, documents, databases, APIs | API changes on every connected system |
| Document understandingExtraction from invoices, emails, PDFs, scans | Model updates, new document types |
| Exception handling with peoplePause, ask, review queues, guidance, resume from the paused step | Workbench operations, reviewer tooling |
| Model management and evaluationsModel routing, prompt and model versions, regression and drift checks | Model upgrades, evaluation suites, inference cost control |
| Authoring, versioning, and the logic runtimeWriting logic in English, drafts, publish, rollback, the interpreter | Language and runtime evolution |
Your options
Three ways to get business logic into production
03Every company picks one, on purpose or not. Only one keeps your builders on the logic without starting a platform team.
| Buy SaaSThe vendor’s logic, in the vendor’s menus | Build everythingYour logic, on a platform you build and run | Our positionBuild the logic on a platformYour logic, in plain English, on guarantees you buy | |
|---|---|---|---|
| You own the logic | No | Yes | Yes |
| The business can read it | Barely | Barely | Yes |
| Platform guarantees included | Yes | No | Yes |
| Small team to run it | Mostly | No | Yes |
| Fast to first value | Yes | Mostly | Yes |
| Light in year two | Mostly | No | Yes |
| The catch | Your process bends to the product. | A platform company you didn’t mean to start. | Test that the platform doesn’t quietly constrain the logic. |
The honest exception
Should you build the platform too?
Sometimes. Only if all four answers are yes.
- Is the platform itself your product?
- Do you already run a platform team with 24/7 on-call?
- Can you pass SOC 2 and customer security reviews on it this year?
- Will you fund the team that runs it for as long as the logic runs?
Your answer
Answer the four questions.
Building the platform only makes sense if every answer is yes.
Works with Microsoft
Already on Copilot Studio or Power Automate? Keep them.
04Let Copilot handle the conversation. Let Kognitos run the rules the same way every time, with a record.
- AskSomeone asks in Copilot.
- UnderstandA Copilot Studio agent reads the question and calls the automation.
- RunKognitos runs the rule, written in English as Code, the same way every time.
- RecordThe answer comes back with every decision and its reason, replayable.
Power Automate
Best for routine glue across Microsoft 365: approvals, notifications. Strains on exception-heavy, auditable work, where the logic hides in boxes and expressions.
Copilot Studio
Best for conversation and fast agents grounded in Microsoft 365. Strains on rules that must always hold, because a model plans the control flow at run time.
Kognitos
Best for rules that must run the same way every time, exceptions resolved by people, audit and replay. Works with Copilot Studio, which can call Kognitos automations.
Resources
Go deeper on building and buying
The position paperThe full argument from Binny Gill, Founder & CEO.Read the paper
What is deterministic AI?The model reads, the rules act, and a person decides the doubtful cases.Learn more
English as CodeBusiness logic the business can read, that runs the same way every time.Learn more
AR automation: build, buy, or agentic?The same decision, worked through for accounts receivable.Learn more
FAQs
Questions about building and buying
Is SaaS dead?
Not as software people buy. What is dying is the bundle: the assumption that to get a reliable platform you must also accept a vendor's version of your business logic. AI makes the logic cheap to build, so companies will own more of it and buy the platform underneath.
Should we build or buy AI agents?
Build the business logic, because it encodes how your company works and should be yours. Buy the platform it runs on: deployment, scaling, security, governance, audit, monitoring, and SLAs. Avoid point solutions that encode your logic in their own menus.
Can coding agents build our platform?
They can write much of the code, quickly. They can't carry the on-call rotation, sit through the SOC 2 audit, patch vulnerabilities, keep connectors working as APIs change, or explain old decisions to an auditor. Those recurring costs are people costs, not token costs.
What should we never build ourselves?
Anything that isn't your business and has to be correct every day: the runtime, scaling, security, compliance evidence, audit and replay, and connectors to systems that change underneath you.
We already have Copilot Studio and Power Automate. Do we need another platform?
They are good at conversation, Microsoft 365 grounding, and routine flows. Business logic that must run the same way every time, with exceptions handled by people and a replayable record, needs a platform built for it. They work together: Kognitos automations can be called from Copilot Studio.
What is a platform that leaves the business logic unspecified?
One that specifies everything SaaS used to carry and constrains nothing about your logic. Any rule can be expressed fully, the business can read it, it runs deterministically where it matters, and exceptions are resolved by people. Kognitos does this with English as Code.
When does building the platform yourself make sense?
When the platform is your product, when you already run a platform team at scale, or when a constraint no vendor can meet forces your hand. Otherwise it means entering a business you didn't plan to be in, for as long as the logic on top of it runs.
How much does it cost to build an AI automation platform?
With coding agents, writing the code is the smaller part. The larger, recurring cost is supporting every platform guarantee in-house: on-call, security and compliance, connectors, and model operations, for as long as the logic runs.
Own the logic. Buy the guarantees.
The full argument: what SaaS really sold, what AI changes and what it doesn’t, and the questions to ask before you build.