# Manisha Parwani > Manisha Parwani is an AI engineer in Helsinki who builds agents you can trace: Claude and Azure AI systems shipped with evals, guardrails and tracing. - Role: AI Engineer - Location: Helsinki, Finland. EU Blue Card holder, based in Finland. - Availability: Open to AI engineering roles in Helsinki and remote across the EU. - Email: hi@manisha.dev - LinkedIn: https://www.linkedin.com/in/manisha-parwani-447b491aa ## Pages - [Home](https://manisha.dev/): who I am, my career as a trace, skills, principles, FAQ, contact - [Work](https://manisha.dev/work/): case studies and lab builds - [About](https://manisha.dev/about/): my story, certifications, recognition, languages - [CV](https://manisha.dev/cv/): full CV, also as [PDF](https://manisha.dev/manisha-parwani-cv.pdf) ## Work - [A second opinion for every AI decision](https://manisha.dev/work/second-opinion-agent/) (Sanoma Media Finland, 2025): AI validation of advertising creatives, plus a Claude agent that reviews the AI's verdicts so human reviewers know when to look twice. - [From booking to live, without the spreadsheet](https://manisha.dev/work/booking-to-live-pipeline/) (Sanoma Media Finland, 2025): Orchestrating ad creatives across a CRM, two ad servers and AI services, deployed with infrastructure as code and passwordless CI/CD. - [Keeping a global corporate site steady](https://manisha.dev/work/avanade-sitecore/) (Accenture for Avanade, 2021): Design, build and regression testing on avanade.com, with integrations across search, video and marketing platforms, released monthly. - [An agent that reads the logs first](https://manisha.dev/work/root-cause-agent/) (lab build, 2026): An MCP server that lets Claude query Application Insights and Log Analytics, propose a root cause for an incident, and show the queries behind every claim. - [Runbooks that answer back, with a scorecard](https://manisha.dev/work/runbook-rag-evals/) (lab build, 2026): Retrieval over runbooks and architecture decision records, graded by golden questions and an LLM judge on every pull request. ## Experience ### Software Engineer, Sanoma Media Finland (Helsinki, Mar 2025 to now) - Built full-stack systems end to end with .NET, Vue (TypeScript) and Azure, from an AdOps review dashboard to backend APIs and Durable Functions pipelines. - Integrated Google Ad Manager, Xandr, CRM and Azure AI services to automate the ad creative lifecycle from booking to live delivery. - Built AI validation of ad creatives with Azure AI Content Understanding, and a Claude Agent SDK agent that gives AdOps reviewers a second opinion on AI decisions. - Secured APIs with Entra ID and role-based authorization, and set up passwordless CI/CD (OIDC, managed identities) with least-privilege RBAC. - Wrote Bicep, GitHub Actions pipelines, architecture decision records and release checklists, and delivered a new service to production for the first time. - Added retry logic, OpenTelemetry tracing and automated tests (xUnit, Vitest, Playwright), and traced production issues to root cause. ### Senior Software Analyst, then Development Team Lead, Accenture (Mumbai, May 2021 to Sep 2024) - Led a team of engineers delivering Sitecore solutions, using Scrum to analyse requirements and prioritise work. - Enhanced avanade.com, the Sitecore-based corporate site, through design, implementation and regression testing. - Integrated Brightcove and Coveo with Sitecore and marketing tools (Adobe, DemandBase, Marketo), working directly with vendors. - Ran monthly automated releases through Azure DevOps, with post-deployment monitoring and validation. ### Senior Software Developer, Systenics Solutions (Mumbai, Dec 2016 to Apr 2021) - Delivered e-commerce applications across the full lifecycle, improving performance through storage and processing-time optimisation. - Ran three projects at once with a team of three developers, gathering requirements with product owners and stakeholders. - Monitored development, test and staging environments and managed production deployments to Azure. ## Skills - Agents and LLM apps: Claude Agent SDK, Claude Code and Skills, Model Context Protocol servers, Tool use and structured outputs, Microsoft Agent Framework, Semantic Kernel, Azure OpenAI, Microsoft Foundry, Multi-agent orchestration, LangGraph, Prompt engineering - Retrieval and documents: RAG pipelines, Embeddings, Azure AI Search, pgvector, Chunking and reranking, Azure AI Content Understanding, Document intelligence - Evals and safety: Golden datasets, LLM-as-judge, promptfoo, Human-in-the-loop review, Guardrails, Azure AI Content Safety, Prompt-injection defence - LLMOps and observability: OpenTelemetry GenAI tracing, Langfuse, Application Insights, Log Analytics and KQL, Cost and latency budgets, Model routing, Evals in CI - Platform: C# and .NET, .NET Aspire, TypeScript and Vue, Python, Azure Functions and Durable Functions, Container Apps, Service Bus, API Management, Front Door, Bicep, GitHub Actions, Azure DevOps, Entra ID, RBAC, OIDC, SQL Server, MySQL, Oracle - How I work: Root-cause analysis, Production support, Release management, Architecture decision records, Requirements analysis, Scrum, Team leadership, Vendor coordination ## How I work - Trace every decision: A model call is a span: input, output, cost, latency and the reason. If I cannot replay why an agent did something, it is not ready. - Give the model a second opinion: Agents propose, people decide, and the disagreements are the most useful data in the system. - Evals before opinions: A golden set and a judge in CI settle arguments about prompts faster than any meeting. - Own it to production: Infrastructure as code, least-privilege identity, a release checklist and a runbook. The demo is day one. ## Certifications - Azure Developer Associate (AZ-204), Microsoft - Azure Fundamentals (AZ-900), Microsoft - Power Platform Fundamentals (PL-900), Microsoft - Scrum Fundamentals Certified (SFC), SCRUMstudy - People Leadership Credential (PLC) ## Education and languages - Bachelor of Engineering, Computer Science, Vivekanand Education Society's Institute of Technology (VESIT), University of Mumbai, 2012 to 2016 - English: Full professional - Hindi: Native - Finnish: Elementary, learning - German: Elementary ## FAQ ### Who is Manisha Parwani? Manisha Parwani is an AI engineer based in Helsinki, Finland. She works at Sanoma Media Finland, where she builds AI validation of advertising creatives with Azure AI and a Claude Agent SDK agent that gives human reviewers a second opinion. Before that she spent eight years delivering enterprise .NET and Azure systems in Mumbai, including leading a Sitecore team at Accenture. ### What does Manisha build? Agents you can trace: LLM applications and tool-using agents on Claude and Azure AI that ship with evals, guardrails and OpenTelemetry tracing, deployed with infrastructure as code and least-privilege identity. ### Which AI tools and platforms does she work with? Claude Agent SDK, Claude Code, Model Context Protocol, Azure OpenAI, Microsoft Foundry, Azure AI Content Understanding, Azure AI Search, Semantic Kernel and Microsoft Agent Framework, with promptfoo and Langfuse for evals and tracing. The platform underneath is .NET, TypeScript, Python, Azure Functions, Container Apps, Bicep and GitHub Actions. ### Is Manisha open to new roles? Yes. She is open to AI engineering roles in Helsinki and remote across the EU. She holds an EU Blue Card and lives in Finland. ### How do I contact Manisha? Email hi@manisha.dev or use the contact form on manisha.dev. She also answers on LinkedIn.