Farzan Ansari
Farzan owns technical direction, architecture, integrations and deployment at Kastling. He works across AWS and Azure, and designs the approval points that keep a named person in control of consequential decisions.
LinkedInWritten by Farzan
Monitoring AI in production: what to log, measure and alert onMonitoring AI in production means watching four things that ordinary uptime checks miss: whether the answers are still good, what each task costs, how the system behaves when inputs change, and whether the human approval step is keeping up. Log every model call and tool action with sensitive data protected, sample real outputs for review, alert on thresholds a named owner agreed in advance, and write down how to roll back.Read RAG, fine-tuning or prompting: how to give AI your company's knowledgeUse retrieval (RAG) when the AI must answer from facts that change and show where each answer came from. Use fine-tuning when you need the model to behave consistently, for example a fixed output format or a narrow classification. Start with prompting, because a small, stable body of context often needs neither.Read How to write an AI acceptable use policy for employeesAn AI acceptable use policy tells employees which AI tools they may use, what data may go into them, how closely to check the output, and who to ask when a new tool shows up. It matters now because most employees already use AI on their own initiative, often through personal accounts the company cannot see. A workable policy fits on two pages, matches its rules to how sensitive the data is, and names one owner who keeps it current.Read AI agents with human approval: how to keep a person in controlAI agents with human approval means deciding in advance which actions a person must see before they happen, and how much of the work still reaches them. Money, legal terms, customer messages and irreversible changes need a person by default, while routine, reversible work can run on sampled review instead. Get the design wrong and you either slow the agent to a crawl or approve items nobody actually checked.Read AI agents, RPA or workflow automation: which fits the work?Workflow automation connects systems through APIs, RPA drives an existing screen when no API exists, and an AI agent interprets a request and chooses its own steps. Pick between them based on what the systems allow and how much of the work follows a written rule, not on which sounds newest. Most working processes end up combining a workflow, an AI step and a person approving anything consequential, rather than relying on just one.Read Why AI pilots stall, and how to get one into productionAI pilots usually stall for operational reasons that were visible from day one: no named owner, a demo run on clean data, and no plan for the exceptions and approvals real work produces. This piece covers why that gap opens, seven decisions that let a pilot become production without a second project, and the explicit criteria for deciding whether to scale, fix or stop. It also covers what changes, in monitoring, access, support and cost, once a pilot goes live.Read Data security and governance for AI in the enterpriseAI data security and governance means knowing which data each AI workflow touches, approving the sources it may use, checking what your AI provider does with prompts and outputs, and making the AI respect the same permissions as the person using it. It also covers defenses against prompt injection, logging with sensible retention, human approval for consequential actions and a clear usage policy for staff. Frameworks such as the NIST AI RMF and ISO/IEC 42001 give that work a structure.Read AI document processing for invoices, contracts and formsAI document processing combines three techniques: OCR to turn images into text, layout models to understand where values sit on a page, and language models to extract and interpret fields that vary between documents. The technique matters less than what surrounds it: validation rules that check extracted values against your own records, confidence thresholds that decide what a person sees, and a review queue somebody owns. Measure field level accuracy and the share of documents that pass without a human touch, not a single headline accuracy number.Read Running AI in production on AWS and AzureRunning AI in production on AWS or Azure means giving a model the same controls as any business system: identity, secrets, logging, cost limits, fallbacks and a named owner for every operating task. Managed services such as Amazon Bedrock and Azure OpenAI in Microsoft Foundry host the models, but the architecture, data boundaries and day-to-day operations remain your decisions to make and assign.Read Building an AI assistant on your company knowledgeAn AI knowledge assistant answers questions using your own approved documents rather than general model training, a pattern called retrieval-augmented generation. The work is less about the model and more about choosing which sources count as authoritative, enforcing the permissions those sources already carry, showing citations people can check, and deciding what happens when information is missing or two documents disagree. Buy Microsoft 365 Copilot, Gemini for Workspace or Glean when your knowledge lives in one mainstream suite, and build when it is spread across line of business systems or the answers must trigger work.Read AWS or Azure for AI workloads: how to chooseFor most companies, the right cloud for AI is the one where their identity, data and technical team already sit. As of September 2026, Amazon Bedrock and Microsoft Foundry both offer models from OpenAI, Anthropic and other providers, so model choice rarely settles the question alone. Identity, data location, compliance scope, skills and existing agreements usually do.Read Connecting AI to your ERP, CRM and data warehouseConnect AI to your ERP, CRM and data warehouse through the vendor's supported APIs, webhooks, event streams or an integration platform, not by writing straight to the database. Start with read access, send every write through a named approver until the workflow has proven itself, and check API limits, licenses, permissions and data quality before you commit to a design.Read Planning a legacy application migration to the cloudA legacy application migration to the cloud starts with deciding, application by application, whether to retire, retain, rehost, replatform, refactor or replace it. Then map dependencies and data, set a cost baseline, plan cutover with a tested rollback, and treat security and identity as part of the move rather than an afterthought. Add AI or automation during the migration only where you are already changing that part of the system.Read Model-agnostic AI architecture: avoiding vendor lock-inA model-agnostic AI architecture puts one internal interface between your application and any model provider, so swapping models is a configuration change rather than a rewrite. It matters because models are retired on published schedules, often within 12 to 18 months of launch. The real lock-in usually sits elsewhere: in proprietary vector stores, agent frameworks, fine-tunes and contract commitments.Read Technical debt in business systems: how to spot it and pay it downTechnical debt is the extra effort every future change costs because of shortcuts, age and neglect in the systems you already run. Leaders see it as slow changes, fragile integrations, spreadsheets holding processes together, one person nobody can replace, and software versions the vendor no longer supports. Pay it down where it blocks work you actually plan to do, in slices, rather than through a full rewrite.Read How to test AI before it touches your operationsTesting AI before deployment means scoring it against a set of real past cases, with the business owner agreeing in advance what counts as correct and what target has to be met. Run it in shadow mode or a limited pilot before it can change anything, keep human review on consequential decisions, and rerun the same test set every time a prompt or model changes.Read
See how an idea becomes a working solution.
Explore how we scope, build and hand over AI projects, with clear decisions at each stage.