Frequently asked questions.

Grids and Guides Technologies is a Data Science, Machine Learning and AI company that helps organisations solve complex business and operational problems using data and intelligent technology.

Our work spans the full spectrum from Data Science and Machine Learning to fine tuning, SLMs, LLMs, AI agents, semantic systems, ontologies and knowledge driven AI architectures. Where required, we also build the applications and technology infrastructure needed to put these capabilities into real business use.

Our primary focus is Data Science, Machine Learning and AI.

However, we also provide application development and support services to some of our long standing customers. These engagements are generally enterprise grade systems where we have an established relationship and a deeper understanding of their technology environment.

We do not position ourselves as a general purpose website, ecommerce or small application development company.

We don't see AI as simply RAG + LLM + chatbot + agent.

Our work can extend from data and statistical modelling through Machine Learning and model fine tuning to SLMs, LLMs, semantic layers, ontologies, knowledge graphs and autonomous systems.

More importantly, we start with the business problem and determine what technology is actually required. Sometimes the right answer is an LLM. Sometimes it is traditional Machine Learning. Sometimes it is a combination of models, structured data, semantic representations and enterprise systems.

The technology is a means to solve the problem, not the product we are trying to sell.

We are not limited to one of those categories.

Grids and Guides operates across consulting, Data Science, Machine Learning, AI and enterprise technology delivery.

We can help an organisation identify where AI makes sense, design the technical approach, build the models and intelligent systems, integrate them with existing technology and support the resulting enterprise application.

That allows us to stay involved from problem definition through implementation, rather than handing a strategy to one vendor and development to another.

Our engagements involve the relevant technical leadership, Data Science, Machine Learning, engineering and domain expertise required for the problem.

For suitable engagements, customers can also work directly with senior members of the organisation rather than having all communication filtered through a conventional sales layer.

No.

RAG, LLM applications and AI agents are only part of what we do.

Our work can extend into Machine Learning, model fine tuning, Small Language Models, Large Language Models, semantic layers, ontologies, knowledge graphs, reasoning systems and autonomous agents.

The architecture is determined by the problem. We do not force every business problem into a RAG chatbot or an AI agent simply because it is currently popular.

Yes.

We work across different levels of the AI stack, including fine tuning existing models, adapting models for specific use cases, building and deploying SLM based solutions, working with LLMs and integrating models into larger enterprise systems.

The appropriate approach depends on the data, use case, performance requirements, cost, privacy and deployment environment.

Yes.

For complex enterprise problems, we can work with ontology engineering, semantic layers, knowledge graphs, knowledge representation, contextual reasoning and autonomous agent architectures.

These approaches become particularly relevant when an organisation needs AI to understand relationships, business concepts, rules and context rather than simply retrieve information or generate text.

We work with structured, semi structured and unstructured data.

This can include databases, transactional data, documents, text, machine and sensor data, maintenance records, logs, APIs, application data, images, audio and other enterprise data sources.

The objective is not simply to process data, but to determine how that data can be used to solve the underlying business problem.

We are not tied to a single model, framework or technology provider.

Depending on the problem, we can work with commercial LLMs, open source models, SLMs, traditional Machine Learning models and custom or fine tuned models.

Technology selection is driven by requirements such as accuracy, latency, privacy, cost, data availability, deployment constraints and maintainability.

Yes.

Where required, solutions can be designed for private cloud, on premise or controlled enterprise environments, subject to the customer's infrastructure and security requirements.

The deployment architecture is determined as part of the technical design.

Yes.

We can work with existing databases, APIs, enterprise applications, cloud environments, legacy systems, production systems and other technology already used by the organisation.

Where necessary, we build the application, integration and engineering layers required to make the AI or ML capability usable within the existing environment.

Yes.

Consulting can involve identifying suitable AI use cases, evaluating existing systems, defining an AI strategy, assessing technical feasibility, designing the solution architecture and determining whether AI is actually appropriate for the problem.

We believe the technology should follow the problem, not the other way around.

Yes, when the engagement requires it.

We have long standing customers for whom we provide application development and ongoing support in addition to our AI, ML and Data Science work.

However, our application development focus is primarily enterprise grade systems, not general purpose websites, ecommerce stores or small applications.

Yes.

We can assess and take over existing projects that require technical recovery, restructuring, completion, modernisation or further development.

This includes situations where another technology partner has started a project but the expected outcome has not been achieved.

Our current areas of focus are Manufacturing, Supply Chain, Education, Finance and Non-Profit.

These are our focus areas, not restrictions. We are open to working with organisations in other industries when the problem is a strong fit for our capabilities and the customer can provide the necessary domain knowledge or subject matter expertise.

Our approach is to combine our technical expertise with the customer's domain expertise rather than pretend that every industry can be understood without context.

Yes. Depending on the engagement, we work with domain subject matter experts alongside our Data Science, Machine Learning and engineering teams.

Domain knowledge is particularly important for complex areas such as manufacturing, supply chain and finance, where understanding the business process is often as important as building the model itself.

Yes.

Our focus industries give us accumulated experience and domain understanding, but they are not a closed list.

If a customer has strong domain expertise and is willing to work collaboratively with our team to understand the business context, we are open to applying our technical capabilities to other industries as well.

No.

Our capabilities can be applied across enterprises, institutions, startups and other organisations where there is a meaningful problem that requires Data Science, Machine Learning or AI.

The suitability of an engagement depends more on the problem and the organisation's commitment to solving it properly than on its organisational category.

We can work with both.

The deciding factor is not simply the size of the organisation. We look at the complexity of the problem, the seriousness of the engagement, the technical requirements and the customer's approach to quality and long term partnership.

No.

We are selective about the organisations and engagements we take on.

We prefer working with customers who value quality, technical depth and long term partnership rather than selecting a vendor primarily on the lowest cost.

A strong working relationship and alignment on expectations are important to us.

Our MVP engagements typically start at $30,000.

This is for a scalable, working Agentic AI MVP designed around a real business use case, not a generic demo or a quick prototype generated using Claude Code or similar tools.

The actual investment depends on the complexity of the problem, integrations, data requirements, architecture, and level of intelligence required.

Our consulting rates depend on the expertise required and the nature of the engagement:

• Technology Consulting with Vasanth: starts at $50/hour
• AI Strategy and Use Case Identification with Kaarthick: starts at $100/hour
• Domain SME Consulting with Ontology and Domain Modelling: $450/hour

For specialised domains such as manufacturing, our experts spend time understanding the actual business and operational problem, studying the domain context, and contributing to the solution through structured domain knowledge, ontology, process understanding, and technical direction.

It is not simply a consulting call or a conversation. The value comes from the expertise brought into the problem. Once you work with us, you will understand the difference.

Our resource rates start at $15/hour for entry-level resources and can go up to $100/hour, depending on the individual's expertise, technology specialisation, experience, and the value they bring to the engagement.

We can provide resources across different levels of technical and domain expertise based on the requirements of the project.

Our pricing is transparent and consistent.

We maintain standard rates across customers rather than quoting one price to one customer and a different price to another for the same level of expertise.

This is our standard MRP. The rate reflects the expertise, experience, and value associated with the person or service involved.

For project engagements, the final commercial proposal can vary based on the scope, duration, complexity, and team composition, but our underlying rates remain transparent.

There is no single price for an AI project because the cost depends on the problem, data, model requirements, integrations, infrastructure and deployment environment.

As a general reference, an engineered AI MVP typically starts around $30,000. Larger enterprise implementations are scoped separately after understanding the requirements.

Timelines depend on the complexity of the problem, data readiness, integrations, model development and validation requirements.

A properly scoped AI MVP can typically take several months, while larger enterprise implementations can require a longer phased engagement.

We provide a timeline after understanding the actual problem and prerequisites rather than giving an arbitrary fixed duration upfront.

We generally do not undertake unpaid Proofs of Concept.

Before development begins, we prefer to understand the problem, establish the requirements, assess the available data and define the scope properly.

Where a POC is appropriate, it is treated as a properly scoped engagement rather than an unpaid experiment.

We treat customer information and project details as confidential and do not disclose project details simply for marketing purposes.

Specific confidentiality, data handling, security and intellectual property requirements are established as part of the engagement and contractual arrangements.