1. About Flentas Technologies
Flentas Technologies is a Pune-headquartered cloud consulting and delivery company with offices in Mumbai, Hyderabad and Dubai. An AWS Advanced Consulting Partner holding the AWS Generative AI, DevOps and IoT Competencies and a 2025 AWS Partner of the Year award recipient, Flentas delivers cloud migration and modernisation, DevOps, managed services, cloud security, data engineering, IoT and Generative AI solutions to enterprises across India and the GCC through a team of more than 100 certified cloud professionals, with particular depth in banking, financial services, insurance and other regulated industries.
2. Position Summary
The AI Solutions Architect is the senior technical owner of Flentas Technologies' Generative AI and Agentic AI engagements. The role spans the full engagement lifecycle: qualifying and shaping opportunities with the sales team and AWS, leading technical discovery, designing the target architecture, authoring and pricing the Statement of Work, and remaining the accountable architect through delivery to production.
The role also carries responsibility for the Generative AI offerings catalogue and sales collateral, the internal AI-driven development toolchain, AWS partner programmes for AI, and the mentoring of Generative AI engineers. Approximately half of the role is presales and solution design; the remainder is architecture ownership during delivery, practice development and team enablement. The position works closely with the Data Solutions Architect on engagements that combine data platforms and AI.
3. Key Responsibilities
3.1 Presales and Solution Design
- Lead technical discovery for Generative AI and agentic opportunities, including discovery calls and workshops, mapping of business processes to prioritised AI use cases, and preparation of use-case documents, assessment reports and proposals (including AWS-template AI assessment reports with ARR and ROI estimates).
- Author Statements of Work and proposals covering scope, target-state architecture, phased plans (proof of concept, pilot and production), acceptance criteria (accuracy, precision and recall, latency and cost per unit), assumptions, risks, team composition and timelines.
- Prepare commercial and technical estimates: AWS Pricing Calculator estimates, LLM cost models per document, alert or conversation, token-consumption forecasts, and cloud versus on-premises GPU total-cost-of-ownership comparisons where required.
- Produce architecture diagrams and solution presentations for executive, technical and legal or compliance audiences.
- Respond to AWS-sourced inbound opportunities and co-sell requests, and represent Flentas technically in AWS account-team and partner discussions.
3.2 Architecture and Delivery Ownership
- Design production-grade Generative AI systems on AWS using Amazon Bedrock (Anthropic Claude, Amazon Nova, OpenAI models on Bedrock and other model families), Bedrock Agents and AgentCore, Bedrock Knowledge Bases, Guardrails and cross-region inference profiles, and Amazon SageMaker or vLLM where self-hosted or open-weight models are required.
- Design agentic systems: multi-agent orchestration (Strands Agents, LangGraph, CrewAI, n8n, AWS Step Functions and EventBridge), tool use and Model Context Protocol (MCP) servers, human-in-the-loop approval gates, autonomy thresholds, shadow-mode rollouts and evaluation harnesses.
- Design retrieval-augmented generation (RAG) and knowledge layers: chunking and embedding strategy, vector stores (Amazon OpenSearch Serverless, Aurora PostgreSQL with pgvector), retrieval services, and de-identified ingestion pipelines (Amazon Textract, Bedrock Data Automation, Amazon Comprehend, Presidio, Amazon Macie) orchestrated with Step Functions.
- Design for regulated industries: region and data-residency selection (ap-south-1, me-central-1, EU regions), DPDP-, DIFC- and GDPR-aligned controls, PII tokenisation, AI gateways (for example LiteLLM on ECS Fargate), AWS CloudTrail, KMS and Secrets Manager, and model-provider data-handling terms; respond in writing to client legal and compliance queries.
- Conduct model selection and evaluation: build adjudicated evaluation sets, define precision, recall, latency and cost thresholds, recommend model tiers per task with a defensible cost rationale.
- Act as the accountable architect through delivery: technical lead for the engagement team (team lead, GenAI, backend and frontend engineers and QA), design reviews, milestone gates, acceptance sign-off and client technical escalations.
3.3 Practice Development and AWS Partnership
- Own and maintain the Generative AI offerings catalogue, the Enterprise Agentic AI Strategy sales deck and the AI use-case library.
- Own the AI-driven development lifecycle (AI-DLC) toolchain and internal AI intellectual property, including adoption of Kiro and Claude Code on Amazon Bedrock, MCP-based accelerators such as ChangeSafe AI, packaged workflow rules, enterprise-plan evaluation and model-access governance for the delivery organisation.
- Drive AWS partner programmes for AI: ACE opportunity registration, AWS funding programmes (AI assessment, proof-of-concept and migration funding), Generative AI Competency evidence and joint go-to-market activity with AWS account teams.
- Contribute reference architectures, presales accelerators (discovery cards, estimation templates) and documentation to the Cloud Center of Excellence (Confluence).
3.4 Team Development and Thought Leadership
- Mentor Generative AI engineers; run internal enablement (training modules and certification tracker) for the AI practice; participate in the hiring of AI engineers.
- Represent Flentas externally through AWS Community Day sessions, technical blogs, customer webinars and partner events.
4. Required Qualifications and Experience
- Bachelor's degree in Computer Science, Engineering or a related discipline.
- 6–10 years of experience in software, cloud or solution engineering or architecture, including a minimum of 2 years designing and delivering LLM-based systems in production (RAG, agents, intelligent document processing or conversational AI).
- Demonstrated presales experience: technical discovery, use-case prioritisation, SOW and proposal authoring, cost estimation and architecture presentations to executive and technical audiences.
- Experience with regulated-industry requirements in BFSI, insurance or financial services, including PII handling, data residency and DPDP; exposure to DIFC or GDPR is an advantage.
- Excellent written and verbal English; SOWs, assessment reports and client correspondence are core outputs of the role.
5. Required Technical Skills
- Amazon Bedrock (models, agents, knowledge bases, guardrails, inference profiles) and the surrounding AWS services: Lambda, API Gateway, ECS / Fargate, Step Functions, EventBridge, DynamoDB, Aurora, S3, OpenSearch, IAM, KMS, Secrets Manager and CloudTrail; proficiency with the AWS console, CLI and infrastructure as code (CDK or Terraform).
- Agentic AI: tool use / function calling, MCP, at least one orchestration framework (Strands Agents, LangGraph, CrewAI or equivalent), evaluation, guardrails and human-in-the-loop patterns.
- Strong Python; ability to read TypeScript / JavaScript and review engineering team code.
- LLM economics: tokens, model tiers, prompt caching, cross-region inference and batch versus real-time inference, with the ability to defend estimates to finance stakeholders.
6. Preferred Skills
- Anthropic ecosystem (Claude Code, Claude Desktop, MCP servers), Kiro or Amazon Q Developer, and AI-DLC methodology.
- Multimodal and video-language models (Amazon Nova, VLMs), self-hosting with vLLM, GPU sizing and total cost of ownership.
- AWS partner-ecosystem experience: ACE, funding programmes and competency submissions. (Good to have)
- Amazon SageMaker, MLflow, and classical machine learning or computer-vision fundamentals.
- Prior experience leading a 4–8 person delivery team as team lead or architect.
7. Certifications (Good to Have)
Certifications are desirable but not mandatory; demonstrated hands-on delivery experience carries greater weight in selection.
- AWS Certified Solutions Architect – Associate or Professional.
- AWS Certified Machine Learning – Specialty,
- AWS Certified AI Practitioner or AWS Certified Generative AI Developer – Professional (or equivalent).
8. Key Working Relationships
- Internal: Chief Technology Officer and co-founders, Data Solutions Architect, Senior Solutions Architects, account managers, delivery team leads and engineers.
- External: AWS partner-development and account teams; client CXOs, product heads and legal or compliance stakeholders in India and the GCC.
9. Professional Competencies
- Consultative, outcome-oriented engagement with senior client stakeholders.
- Commercial acumen: the ability to scope, estimate, price and defend an engagement.
- Ownership and accountability across the presales-to-delivery lifecycle.
- Structured written communication and documentation discipline.
- Mentoring and collaborative leadership within multidisciplinary teams.
