Website:
sarklabs.com
Job details:
About Sark Labs
Sark Labs is a fast-growing brain training and education company building technology-enabled products for intelligent reading, memory, focus, thinking, creativity and productivity founded by Rajneesh Barapatre who is the World Champion. We use the Zoho ecosystem extensively across our customer operations, learning delivery, CRM, analytics, automation and internal business processes. We already have an active Zoho Creator application and plan to build multiple additional applications. This is an internal product ownership role - not a client-service or ticket-based support role. You will work closely with the founders & co-founders, who have a software engineering background, and will have significant ownership over architecture, development standards and product delivery.
About the role
- We are building a suite of business and learning applications using Zoho Creator and the wider Zoho ecosystem.
- Our current application includes multiple independently deployed Zoho Creator widgets built using JavaScript, HTML, Deluge, Python and CSS. We are also planning additional applications and integrations across Zoho Creator, Zoho CRM and other Zoho products.
- We are looking for a strong, hands-on Zoho Creator Developer who can understand business requirements, design maintainable solutions, build production-ready applications and take ownership from requirement clarification through deployment.
- The developer will work on custom JavaScript widgets, Deluge functions, APIs, data models, automation, integrations, testing, Git-based version control and production deployments.
- The developer will also be responsible for designing and implementing AI-powered features within our applications. This may include integrations with the OpenAI API, Claude API, Gemini API or other suitable AI platforms.
- The role requires someone who can move beyond basic prompt usage and build secure, reliable and cost-effective AI application workflows connected to Zoho Creator, Zoho CRM, external databases, APIs and business processes.
- The developer should also be comfortable using AI-assisted software development tools such as OpenAI Codex, Claude Code, GitHub Copilot or similar tools while independently reviewing, testing and taking responsibility for the generated code.
What you will own
- Development/Maintenance of custom Zoho Creator widgets using JavaScript, HTML and CSS.
- Deluge functions, workflows, validations, schedules and custom APIs.
- Integrations between Zoho Creator, Zoho CRM and other Zoho applications.
- REST API, OAuth, webhook and third-party service integrations.
- Database design, forms, reports, relationships, permissions and role-based access.
- Git-based source control, code reviews, release documentation and rollback readiness.
- Requirement analysis, solution design, development estimates and technical documentation.
- Debugging production issues and identifying permanent solutions.
- Technical guidance and code reviews for junior developers as the team grows.
- AI integrations across Zoho Creator, Zoho CRM and external applications.
- Development of AI-powered application features using the OpenAI API, Claude API, Gemini API or equivalent platforms.
- Secure server-side handling of AI API credentials, prompts, business rules and sensitive data.
- AI workflow architecture, including model selection, tool calling, structured outputs, retrieval, evaluation, monitoring and fallback handling.
- AI-assisted development workflows using tools such as OpenAI Codex, Claude Code and GitHub Copilot.
- Documentation and version control of prompts, models, AI workflows and evaluation criteria.
Key responsibilities
- Translate business requirements into technical designs within the Zoho ecosystem.
- Build and maintain Zoho Creator forms, reports, pages, workflows, functions and widgets.
- Develop responsive, user-friendly widgets using:
- JavaScript
- HTML5
- CSS3
- Zoho Creator Widget SDK
- jQuery where required by existing code
- Modern browser APIs
4. Write secure and maintainable Deluge scripts for:
- Data validation
- Workflow automation
- Scheduled functions
- Custom APIs
- Record creation and updates
- Cross-application integrations
5. Integrate Zoho applications and third-party systems using:
- REST APIs
- OAuth 2.0
- Webhooks
- JSON
- Zoho Connections
- Custom functions
6. Work with Zoho CRM and, where needed, other Zoho applications such as Zoho Flow, Recruit, Meeting, Analytics, Desk, Books or Billing.
7. Design maintainable data models, relationships, lookup structures and permission rules.
8. Use Git and GitHub for branching, commits, pull requests, reviews, releases and rollback management.
9. Create test cases and verify loading, empty, success, error, permission and refresh scenarios.
10. Protect sensitive information by keeping credentials, tokens and secrets out of frontend widget code.
11. Use AI coding and productivity tools responsibly to accelerate development while personally validating security, accuracy and maintainability.
12. Communicate progress, risks, assumptions and blockers clearly.
13. Design and implement AI-powered features such as:
- Intelligent assistants and chat interfaces.
- Student or user guidance systems.
- Content summarization and classification.
- Personalised recommendations.
- Information extraction from documents or unstructured text.
- Automated analysis and reporting.
- Support-ticket or enquiry classification.
- AI-assisted workflow automation.
- Semantic search and knowledge-base applications.
- AI-generated drafts, insights and next-action recommendations.
14. Integrate applications with AI platforms using secure server-side APIs, webhooks and middleware.
15. Work with relevant AI development concepts, including:
- Large language model APIs.
- Prompt engineering and prompt templates.
- Structured JSON outputs.
- Function calling and tool calling.
- Agentic workflows.
- Embeddings and semantic search.
- Retrieval-augmented generation.
- Conversation and context management.
- Model selection and routing.
- Streaming responses.
- AI evaluation and testing.
- Guardrails and output validation.
16. Design AI workflows that account for:
- Incorrect or hallucinated responses.
- Prompt injection and malicious input.
- Data privacy and sensitive information.
- API failures, timeouts and rate limits.
- Token consumption and operating cost.
- Response time and user experience.
- Logging, monitoring and traceability.
- Human approval for important actions.
- Fallback behaviour when an AI service is unavailable.
17. Ensure that AI-generated outputs are validated before they create, update or delete important business records.
18. Build reusable server-side services or middleware using Deluge, Python or another suitable backend technology when an AI integration should not be implemented directly inside a frontend Zoho Creator widget.
19. Use AI coding agents to accelerate requirement analysis, coding, testing, debugging, documentation and code review without compromising security, maintainability or engineering accountability.
Mandatory skills
- At least 3-7 years of relevant development experience.
- Substantial hands-on experience building applications in Zoho Creator, Deluge, JavaScript, HTML and CSS , AI Tools & API, Git, Github or similar version control
- Experience with Zoho Creator forms, reports, pages, workflows, functions and permissions.
- Experience building or maintaining custom Zoho Creator widgets.
- Experience consuming and building REST APIs.
- Understanding of OAuth, authentication, JSON and webhooks.
- Experience with Zoho apps customization or integration.
- Good understanding of relational databases, SQL and data modelling.
- Ability to debug production issues systematically.
- Ability to understand business requirements and independently propose solutions.
- Clear written and verbal communication.
- Hands-on experience integrating at least one production application with a commercial large language model API such as the OpenAI API, Claude API, Gemini API or an equivalent platform.
- Practical experience building at least one AI-powered application, workflow, feature, prototype or internal business tool.
- Ability to call AI APIs securely from a backend service, Deluge function, Zoho Connection, serverless function or middleware layer.
- Understanding of prompt engineering, structured outputs, function or tool calling and AI response validation.
- Experience parsing, validating and safely processing JSON responses generated by AI models.
- Understanding of model limitations, hallucinations, prompt injection, data privacy and responsible AI usage.
- Ability to protect AI API keys and credentials and avoid exposing them inside frontend JavaScript or Zoho Creator widgets.
- Ability to evaluate AI output quality and create test cases for AI-powered features.
- Practical experience using AI coding assistants such as OpenAI Codex, Claude Code, GitHub Copilot or equivalent tools.
- Ability to review, debug and explain AI-generated code rather than accepting generated code without validation.
Preferred Skills
- Zoho Creator certification and experience across Zoho One, CRM, Flow, Analytics, Books, Billing, Recruit, Meeting and Desk.
- Experience with Zoho Extension Toolkit, widget packaging, deployment, automated testing and CI/CD.
- Backend or middleware development using Python or Node.js for integrations and AI workflows.
- Knowledge of application security, role-based access, data isolation, PII protection and audit logging.
- Practical experience with AI coding assistants and leading LLM platforms, including OpenAI APIs/Responses API, Claude/Claude Code and Gemini.
- Experience with prompt engineering, structured outputs, tool calling, AI agents, multimodal AI and multi-step workflows.
- Experience integrating AI with business applications, CRM systems, databases and workflow platforms.
- Knowledge of embeddings, vector databases, semantic search, RAG and MCP or comparable tool-connection standards.
- Experience building AI solutions for document processing, summarization, classification, recommendations, conversations, speech, text and image understanding.
- Ability to manage AI security, prompt injection, tokens, costs, latency, rate limits, retries and reliability.
Click on Apply to know more.