Skip to content
SP Sarath Panha

Services

Architecture and delivery for AI features that need to work in production.

Pick the engagement that matches where you are: planning, building, or scaling a team that builds.

1 to 2 weeks · fixed price

AI architecture review

An independent assessment of an AI system you run or plan to build, before problems reach your users or your cloud bill.

Good fit if

You have an AI prototype heading to production, or a live feature that is slow, costly or unpredictable.

Pricing

[From $X, fixed price]

What you get

  • Review of architecture, data flow, prompts and model choices
  • Risk register: reliability, security, privacy and prompt injection
  • Cost and latency analysis with concrete savings
  • Target architecture diagram and decision records
  • Prioritised 90-day plan, presented to your team

4 to 10 weeks · fixed scope

AI feature build

I design and build the feature end to end, working inside your codebase and alongside your team.

Good fit if

You know what you want to build and need someone who has shipped it before.

Pricing

[Quoted after scoping]

What you get

  • RAG, agent or LLM integration built to production standard
  • Evaluation suite so quality is measured, not guessed
  • Guardrails, rate limits and cost controls
  • Tracing and dashboards for quality, latency and spend
  • Docs, runbooks and a handover session

Monthly retainer

Fractional architect

Senior architecture leadership a few days a month, without a full-time hire.

Good fit if

Your team is shipping AI features and needs experienced guidance on the hard decisions.

Pricing

[From $X per month]

What you get

  • Architecture and design reviews for new features
  • Decision records (ADRs) and technical roadmap
  • Code reviews on critical paths
  • Mentoring for engineers new to AI systems
  • A standing weekly call and async support

Process

How an engagement works

  1. Step 1

    Discovery call

    30 minutes, free. We talk through the problem, constraints and what success looks like.

  2. Step 2

    Scope and proposal

    A short written proposal: approach, milestones, fixed price or retainer, and risks.

  3. Step 3

    Build with weekly demos

    Working software every week, decisions recorded as ADRs, evals and monitoring from day one.

  4. Step 4

    Handover

    Docs, runbooks and a walkthrough so your team owns it. Optional support after launch.

FAQ

Common questions

Do you work with teams outside your time zone?

Yes. Most work is async with a weekly call at a time that suits both sides. [Confirm your hours and time zone.]

Which LLM providers and clouds do you use?

Whichever fits your constraints. I design so the model provider can be swapped, and work on the major clouds. Each case study lists its exact stack.

Can you work inside our existing codebase?

Yes, that is the usual setup. I follow your conventions, open pull requests, and leave decision records behind.

What happens after handover?

You own everything: code, docs and dashboards. Optional support is available as a small monthly retainer.

Can you sign an NDA?

Yes. Client names in case studies are anonymised unless a client agrees otherwise.

Have an AI feature to design or ship?

Book a free 30-minute discovery call. You will leave with a clear next step, whether or not we work together.

Book a discovery call