DoozieSoft
Policy

Fair AI Usage Policy.

We use AI in our work. We believe that's a strength — when done responsibly. This policy declares how we use AI, what guardrails we enforce, and what our clients can expect.

Our commitment: AI usage will never compromise the quality, security, or confidentiality of what we deliver.

Core Principles

Eight commitments we hold ourselves to

01

AI Augments — It Does Not Replace

We use AI to amplify human capability, not to substitute human judgment. Every AI-assisted output — whether code, documentation, analysis, or client deliverable — is reviewed, validated, and approved by a human before it reaches production or a client.

02

Transparency by Default

If AI was used in any part of delivering a service or building a system, we disclose it. Clients are informed when AI tools assist in code generation, document drafting, data analysis, or system design. No hidden automation. No pretending AI output is purely human work.

03

Quality Is Non-Negotiable

AI tools accelerate our process — they do not lower our standard. Every line of code is reviewed to the same rigour whether it was written by a human or suggested by an AI. Every document is fact-checked. Every system is tested against the same criteria. The use of AI does not reduce the quality of what we deliver.

04

Client Data Never Trains External Models

Client data, business logic, proprietary processes, and confidential information are never used to train, fine-tune, or improve third-party AI models. When we use AI tools internally, we ensure client data is not transmitted to model providers unless explicitly agreed upon in the engagement contract.

05

Human-in-the-Loop for Every Consequential Action

In every system we build that includes AI capabilities, consequential actions — financial transactions, compliance decisions, personnel actions, data deletion — require explicit human approval before execution. This is enforced at the architecture level, not left to process discipline.

06

Full Audit Trail of AI Actions

Every AI action in production systems we build is logged with: the input provided, the output generated, the confidence level, the model used, and the human who approved or overrode the action. These logs are immutable and available for audit at any time.

07

No Autonomous AI in Production Without Guardrails

We do not deploy fully autonomous AI agents in production environments. All AI agents we build operate within defined boundaries — approved action sets, escalation triggers, and mandatory human checkpoints. An agent that cannot explain what it did and why is an agent we will not ship.

08

Bias Awareness and Mitigation

We acknowledge that AI models carry biases from their training data. In systems where AI influences decisions about people — hiring, performance, resource allocation — we implement review mechanisms to detect and mitigate bias. We test for disparate outcomes and document our findings.

Internal Practices

How we use AI in our own work

We practice what we preach. Here's exactly how AI integrates into our engineering and delivery process.

Code Assistance

AI tools assist in code generation, debugging, and refactoring. All AI-suggested code is reviewed, tested, and must pass the same quality checks as human-written code.

Documentation

AI assists in drafting technical documentation, user guides, and internal notes. All documentation is reviewed for accuracy and completeness by the engineering team.

Research & Analysis

AI tools help research technical approaches, summarise documentation, and analyse options. Conclusions and recommendations are always validated by domain experts.

Communication

AI may assist in drafting emails, proposals, and client updates. All external communication is reviewed and sent by humans. No AI-generated message reaches a client without human review.

System Design

AI aids in exploring architecture options and generating initial designs. All system architecture decisions are made by senior engineers and reviewed with clients.

For Our Clients

What this means for you

If you engage DoozieSoft for any service — system build, AI integration, system audit, or ongoing support — you can expect:

check_circle

The quality of our deliverables is the same or better with AI assistance — never lower.

check_circle

Your data will not be used to train any third-party AI model without your explicit written consent.

check_circle

Any AI-generated content in your deliverables will be disclosed upon request.

check_circle

All AI capabilities we build into your systems include human approval, audit trails, and override mechanisms.

check_circle

We will never deploy a fully autonomous AI agent in your production environment without guardrails and your approval.

check_circle

If you prefer zero AI involvement in your engagement, we will honour that. Just tell us.

Accountability

This policy is a living commitment

AI is evolving rapidly. So will this policy. We commit to reviewing and updating it as technology, regulations, and best practices change. But the underlying principle will not change:

“AI is a tool we use responsibly. It does not define our quality. It does not replace our judgment. And it will never compromise what we deliver to the people who trust us with their systems.”

If you have questions about our AI practices — or want to understand how AI is used in your specific engagement — ask us directly. We will answer transparently.

Signed By

Sandhya N

Founder & Proprietrix

Akshay D Joshi

Tech-Head & Co-Founder

Shailesh S Londhe

Sales Head & Co-Founder

DoozieSoft · Bangalore, India · Effective April 2026