2.1 Role Summary
You build the data backbone our engagements run on. That means designing and shipping production ETL and ELT pipelines, standing up governed lakehouse platforms, integrating messy and heterogeneous source systems, and making the resulting data trustworthy enough that both business users and AI systems can rely on it without a human checking every number.
Increasingly, that backbone also serves agentic and retrieval systems — embedding pipelines, vector and search indexes, feature stores, and the clean metadata and access controls that let an agent query enterprise data safely. You do not need to have built agents yourself, but you should be interested in engineering for that consumer, because a growing share of the practice's work depends on it.
This is a hands-on build role on small senior squads. You write the pipelines, own their reliability in production, and present your design directly to client architects. There is no layer of people between you and the client.
2.2 Key Responsibilities
Pipeline & Platform Engineering
Design, build and operate production ETL / ELT pipelines — batch and, where the use case demands it, streaming or near-real-time.
Stand up and evolve cloud lakehouse platforms using layered architectures (raw / conformed / curated, or Bronze / Silver / Gold), making and defending decisions on grain, partitioning, modelling and conformance.
Build curated, reusable data products designed to serve multiple downstream consumers — BI, ML, agentic systems and operational integrations — rather than a single point solution.
Modernise legacy pipelines and warehouses onto current tooling with a defensible migration path, parallel-run strategy and reconciliation approach.
Ingestion & Integration
Integrate heterogeneous sources: relational and NoSQL databases, REST and GraphQL APIs, SaaS connectors, event streams, flat and semi-structured files, and change-data-capture feeds.
Handle real-world source problems — authentication and token refresh, pagination, rate limits, schema drift, late-arriving and out-of-order data, backfills and idempotent reprocessing.
Work with domain-specific and awkward formats where the engagement requires it, including sensor, telemetry and instrumentation data at high frequency, and proprietary exports that need conversion before they are usable.
Data Quality, Observability & Reliability
Author and enforce data-quality rules — completeness, ranges, referential integrity, freshness, distribution shift — with quarantine handling for failing records rather than silent corruption.
Instrument pipelines for observability: run status, volumes, rejects, latency and lineage, with alerting that surfaces failures before the business finds them.
Own reliability in production — on-call or hypercare participation, root-cause analysis, and permanent fixes rather than repeated manual reruns.
Governance, Security & Cataloguing
Implement cataloguing, lineage, role-based access control and data-classification policy appropriate to the client's regulatory posture.
Maintain data dictionaries and source-to-target mappings alongside the BA, so that the platform is comprehensible to people who did not build it.
Apply masking, anonymisation and retention controls where personal or commercially sensitive data is in scope.
Data Foundations for AI & Agentic Systems
Build ingestion and preparation pipelines that feed retrieval systems — document parsing, chunking, metadata enrichment, embedding generation, incremental refresh and deletion propagation.
Provision and maintain vector stores and hybrid search indexes alongside the AI Engineer, with attention to freshness, cost and index rebuild strategy.
Build and maintain feature pipelines and feature stores for ML workloads, keeping training and serving consistent.
Expose governed, well-documented data interfaces that agents and applications can query safely — with the access boundaries an autonomous consumer requires.
Performance, Cost & Delivery
Tune for scale and cost — partitioning, file sizing, caching, compute right-sizing, incremental over full processing. Cloud spend is a design constraint, not an afterthought.
Work in fortnightly sprints, demonstrating working increments on the client's own data. Running pipelines are the deliverable; slides are not.
Produce architecture notes, runbooks and handover documentation as contracted deliverables, and support UAT, go-live and knowledge transfer.
Responsibilities
2.1 Role Summary
You build the data backbone our engagements run on. That means designing and shipping production ETL and ELT pipelines, standing up governed lakehouse platforms, integrating messy and heterogeneous source systems, and making the resulting data trustworthy enough that both business users and AI systems can rely on it without a human checking every number.
Increasingly, that backbone also serves agentic and retrieval systems — embedding pipelines, vector and search indexes, feature stores, and the clean metadata and access controls that let an agent query enterprise data safely. You do not need to have built agents yourself, but you should be interested in engineering for that consumer, because a growing share of the practice's work depends on it.
This is a hands-on build role on small senior squads. You write the pipelines, own their reliability in production, and present your design directly to client architects. There is no layer of people between you and the client.
2.2 Key Responsibilities
Pipeline & Platform Engineering
Design, build and operate production ETL / ELT pipelines — batch and, where the use case demands it, streaming or near-real-time.
Stand up and evolve cloud lakehouse platforms using layered architectures (raw / conformed / curated, or Bronze / Silver / Gold), making and defending decisions on grain, partitioning, modelling and conformance.
Build curated, reusable data products designed to serve multiple downstream consumers — BI, ML, agentic systems and operational integrations — rather than a single point solution.
Modernise legacy pipelines and warehouses onto current tooling with a defensible migration path, parallel-run strategy and reconciliation approach.
Ingestion & Integration
Integrate heterogeneous sources: relational and NoSQL databases, REST and GraphQL APIs, SaaS connectors, event streams, flat and semi-structured files, and change-data-capture feeds.
Handle real-world source problems — authentication and token refresh, pagination, rate limits, schema drift, late-arriving and out-of-order data, backfills and idempotent reprocessing.
Work with domain-specific and awkward formats where the engagement requires it, including sensor, telemetry and instrumentation data at high frequency, and proprietary exports that need conversion before they are usable.
Data Quality, Observability & Reliability
Author and enforce data-quality rules — completeness, ranges, referential integrity, freshness, distribution shift — with quarantine handling for failing records rather than silent corruption.
Instrument pipelines for observability: run status, volumes, rejects, latency and lineage, with alerting that surfaces failures before the business finds them.
Own reliability in production — on-call or hypercare participation, root-cause analysis, and permanent fixes rather than repeated manual reruns.
Governance, Security & Cataloguing
Implement cataloguing, lineage, role-based access control and data-classification policy appropriate to the client's regulatory posture.
Maintain data dictionaries and source-to-target mappings alongside the BA, so that the platform is comprehensible to people who did not build it.
Apply masking, anonymisation and retention controls where personal or commercially sensitive data is in scope.
Data Foundations for AI & Agentic Systems
Build ingestion and preparation pipelines that feed retrieval systems — document parsing, chunking, metadata enrichment, embedding generation, incremental refresh and deletion propagation.
Provision and maintain vector stores and hybrid search indexes alongside the AI Engineer, with attention to freshness, cost and index rebuild strategy.
Build and maintain feature pipelines and feature stores for ML workloads, keeping training and serving consistent.
Expose governed, well-documented data interfaces that agents and applications can query safely — with the access boundaries an autonomous consumer requires.
Performance, Cost & Delivery
Tune for scale and cost — partitioning, file sizing, caching, compute right-sizing, incremental over full processing. Cloud spend is a design constraint, not an afterthought.
Work in fortnightly sprints, demonstrating working increments on the client's own data. Running pipelines are the deliverable; slides are not.
Produce architecture notes, runbooks and handover documentation as contracted deliverables, and support UAT, go-live and knowledge transfer.
Salary : As per industry standard.
Industry :IT-Software / Software Services
Functional Area : IT Software - Application Programming , Maintenance
As an Application Support Engineer, a typical day involves acting as a software detective by investigating and resolving issues across various components of essential business systems. This role requires a proactive approach to identifying challenges and delivering timely solutions to ensure the smooth operation of critical applications. The position demands continuous monitoring, analysis, and collaboration with different teams to maintain system stability and support business continuity effectively.
Roles & Responsibilities:
- Expected to perform independently and become an SME.
- Required active participation/contribution in team discussions.
- Contribute in providing solutions to work related problems.
- Collaborate with cross-functional teams to understand system issues and implement corrective actions.
- Document troubleshooting steps and resolutions to build a knowledge base for future reference.
- Assist in the testing and deployment of system updates and patches to minimize disruptions.
- Support junior team members by sharing knowledge and providing guidance when needed.
Professional & Technical Skills:
- Must To Have Skills: Proficiency in Oracle JD Edwards EnterpriseOne Distribution.
- Strong problem-solving abilities with a focus on diagnosing and resolving software issues.
- Experience in monitoring and maintaining enterprise resource planning systems.
- Ability to analyze system logs and performance metrics to identify root causes.
- Familiarity with business process workflows related to distribution and supply chain management.
- Effective communication skills to liaise with technical teams and business stakeholders.
Additional Information:
- The candidate should have minimum 2 years of experience in Oracle JD Edwards EnterpriseOne Distribution.
- This position is based at our Bengaluru office.
- A 15 years full time education is required
Responsibilities
As an Application Support Engineer, a typical day involves acting as a software detective by investigating and resolving issues across various components of essential business systems. This role requires a proactive approach to identifying challenges and delivering timely solutions to ensure the smooth operation of critical applications. The position demands continuous monitoring, analysis, and collaboration with different teams to maintain system stability and support business continuity effectively.
Roles & Responsibilities:
- Expected to perform independently and become an SME.
- Required active participation/contribution in team discussions.
- Contribute in providing solutions to work related problems.
- Collaborate with cross-functional teams to understand system issues and implement corrective actions.
- Document troubleshooting steps and resolutions to build a knowledge base for future reference.
- Assist in the testing and deployment of system updates and patches to minimize disruptions.
- Support junior team members by sharing knowledge and providing guidance when needed.
Professional & Technical Skills:
- Must To Have Skills: Proficiency in Oracle JD Edwards EnterpriseOne Distribution.
- Strong problem-solving abilities with a focus on diagnosing and resolving software issues.
- Experience in monitoring and maintaining enterprise resource planning systems.
- Ability to analyze system logs and performance metrics to identify root causes.
- Familiarity with business process workflows related to distribution and supply chain management.
- Effective communication skills to liaise with technical teams and business stakeholders.
Additional Information:
- The candidate should have minimum 2 years of experience in Oracle JD Edwards EnterpriseOne Distribution.
- This position is based at our Bengaluru office.
- A 15 years full time education is required
Salary : Rs. 0.0 - Rs. 1,00,000.0
Industry :IT-Software / Software Services
Functional Area : IT Software - Application Programming , Maintenance
Role Overview
We are looking for a skilled Python Developer with strong expertise in Fast API And React to build and manage scalable, high-performance backend systems. The ideal candidate will have hands-on experience in designing microservices, deploying containerized applications, and working in cloud-native environments.
Key Responsibilities
Design, develop, and maintain robust backend services using Python and FastAPI
Build and deploy scalable microservices architectures
Containerize applications using Docker and orchestrate using Kubernetes
Develop and integrate RESTful APIs with high performance and security standards
Collaborate with frontend teams, DevOps, and cross-functional stakeholders
Optimize application performance, reliability, and scalability
Implement CI/CD pipelines and ensure smooth deployment workflows
Monitor, troubleshoot, and debug production issues
Mandatory Skills
Strong proficiency in Python programming
Hands-on experience with FastAPI (or similar modern Python frameworks)
Experience working with Kubernetes for container orchestration
Solid understanding of Microservices Architecture
Experience with Docker and containerization
Experience with Angular/React
Strong knowledge of REST API design and development
Preferred Skills
Experience with cloud platforms such as AWS
Familiarity with CI/CD tools (Jenkins, GitHub Actions, etc.)
Knowledge of databases (PostgreSQL, MongoDB, MySQL)
Exposure to message queues (Kafka, RabbitMQ)
Understanding of monitoring tools like Prometheus, Grafana
Qualifications
Bachelor’s degree in Computer Science, Engineering, or related field
3–6 years of relevant backend development experience
Responsibilities
Role Overview
We are looking for a skilled Python Developer with strong expertise in Fast API And React to build and manage scalable, high-performance backend systems. The ideal candidate will have hands-on experience in designing microservices, deploying containerized applications, and working in cloud-native environments.
Key Responsibilities
Design, develop, and maintain robust backend services using Python and FastAPI
Build and deploy scalable microservices architectures
Containerize applications using Docker and orchestrate using Kubernetes
Develop and integrate RESTful APIs with high performance and security standards
Collaborate with frontend teams, DevOps, and cross-functional stakeholders
Optimize application performance, reliability, and scalability
Implement CI/CD pipelines and ensure smooth deployment workflows
Monitor, troubleshoot, and debug production issues
Mandatory Skills
Strong proficiency in Python programming
Hands-on experience with FastAPI (or similar modern Python frameworks)
Experience working with Kubernetes for container orchestration
Solid understanding of Microservices Architecture
Experience with Docker and containerization
Experience with Angular/React
Strong knowledge of REST API design and development
Preferred Skills
Experience with cloud platforms such as AWS
Familiarity with CI/CD tools (Jenkins, GitHub Actions, etc.)
Knowledge of databases (PostgreSQL, MongoDB, MySQL)
Exposure to message queues (Kafka, RabbitMQ)
Understanding of monitoring tools like Prometheus, Grafana
Qualifications
Bachelor’s degree in Computer Science, Engineering, or related field
3–6 years of relevant backend development experience
Salary : As per industry standard.
Industry :IT-Software / Software Services
Functional Area : IT Software - Application Programming , Maintenance
AI Engineer
Role Summary:
You design and ship the intelligence layer of our engagements. The centre of gravity is agentic AI — systems that plan across multiple steps, call tools and enterprise APIs, retrieve and reason over the client's own data, and complete real work rather than answering a single question. Around that sit the disciplines that make such systems trustworthy: retrieval design, evaluation harnesses, guardrails, observability and human-in-the-loop review.
Applied machine learning remains part of the role. Several engagements need forecasting, classification, anomaly detection or sensor-signal modelling rather than an agent, and the same person is expected to pick the right tool. A candidate who reaches for a language model regardless of the problem is not the profile we are hiring.
This is an engineering role, not a research role. The measure is systems that run in a client's production environment, survive contact with real users and real data, and can be explained to a sceptical business stakeholder — not benchmark scores or prototypes that never left a notebook.
3.2 Key Responsibilities:
Agentic System Design & Engineering
Design and build agentic systems that decompose goals into steps, select and call tools, use enterprise APIs and data sources, and recover sensibly when a step fails.
Define tool and function interfaces an agent can use reliably — clear contracts, validated structured inputs and outputs, sensible error surfaces, and idempotency where actions have side effects.
Implement orchestration and control flow: routing, planning, retries, escalation paths, state and memory management, and deterministic checkpoints where the business cannot tolerate variance.
Design human-in-the-loop and approval boundaries — deciding what an agent may do autonomously, what requires confirmation, and how a human takes over cleanly.
Build multi-agent workflows where they genuinely help, and resist them where a single well-scoped agent or a plain pipeline is the better answer.
Integrate agents with enterprise systems and identity — authentication, authorisation, least-privilege access to data and actions, and full auditability of what the agent did and why.
Retrieval & Knowledge Systems
Build retrieval-augmented systems over enterprise content — documents, tickets, contracts, manuals, wikis and structured data.
Own the retrieval quality chain: parsing and chunking strategy, metadata design, embedding selection, hybrid semantic and keyword search, reranking, and context assembly within practical limits.
Ground outputs in retrieved evidence with citations, and engineer explicitly against hallucination and confident-but-wrong answers.
Work with the Data Engineer on ingestion, incremental refresh, deletion propagation and index freshness so the knowledge layer does not silently drift out of date.
Evaluation, Guardrails & Reliability
Build evaluation harnesses before scaling a system, not after it fails: curated test sets, task-level success criteria, automated regression suites and model-graded evaluation where appropriate.
Define what 'good' means with the client in measurable terms, and report honestly against it — including where the system is weak and where it should not be trusted.
Implement guardrails: input and output validation, prompt-injection and jailbreak defence, PII detection and redaction, content filtering, and refusal behaviour for out-of-scope requests.
Instrument for observability — tracing multi-step runs, capturing tool calls and failures, tracking latency, token consumption and unit cost per task.
Manage cost and latency deliberately: model selection per task, caching, batching, context management, and cheaper fallbacks where full capability is not required.
Applied Machine Learning
Build classical and deep-learning models where they fit the problem better than a language model — forecasting, classification, regression, anomaly detection, clustering, time-series and sensor-signal modelling.
Engineer features from raw operational, transactional or sensor data, and design evaluation that reflects deployment reality — time-aware and grouped splits rather than random ones, and leakage avoidance.
Establish a credible baseline first, then justify added complexity with measured gain.
Work with domain experts to define labels and acceptance criteria in engagements where no ground-truth dataset exists at the outset.
Deployment, LLMOps & Handover
Deploy models and AI services into client environments — containerised services, batch scoring pipelines, or managed cloud AI platforms — with versioning, promotion criteria and rollback.
Monitor systems in production: data and prediction drift, quality regression, cost anomalies and failure patterns; own the retraining or re-tuning cycle.
Close the feedback loop — capture user and expert corrections and route them back into evaluation sets, retrieval improvements, prompt revisions or training data.
Make behaviour explainable to non-technical stakeholders, and deliver model documentation, evaluation reports and operational runbooks at handover.
Client-Facing Solutioning
Shape solutions with clients — assess AI feasibility for a business problem, size effort, identify data gaps, and say clearly when AI is the wrong tool for a given ask.
Demonstrate working increments on the client's own data in fortnightly sprint reviews, and support pre-sales technical discussions and proof-of-value work as required.
3.3 Must-Have Skills & Experience
Area
What We Expect
Experience
5–8 years in machine learning / AI engineering, including meaningful recent experience building LLM-based or agentic systems that reached production. Not research-only or notebook-only.
Core languages
Strong Python — production-quality, version-controlled, tested code. Comfortable building services and APIs (FastAPI or equivalent), not only models.
Agentic engineering
Demonstrable hands-on work with tool / function calling, multi-step agent loops, structured outputs, state and memory management, and failure recovery. Framework experience such as LangGraph, LangChain, LlamaIndex, CrewAI, AutoGen, Semantic Kernel or provider-native agent SDKs — with judgement about when a framework helps and when it gets in the way.
LLM application engineering
Practical depth with frontier model APIs (Anthropic, OpenAI, Google, or open-weight models via vLLM / Ollama / Bedrock): prompt design, structured output enforcement, streaming, context management, caching and cost / latency trade-offs.
Retrieval systems
End-to-end RAG delivery — chunking and metadata strategy, embedding models, vector databases, hybrid search, reranking, grounding and citation. Including knowing why a naive RAG implementation underperforms and how to fix it.
Evaluation
Building evaluation harnesses and regression suites for non-deterministic systems; defining task-level success metrics; model-graded evaluation and its limits; honest error analysis.
Guardrails & safety
Prompt-injection and jailbreak mitigation, input / output validation, PII handling, and designing appropriate autonomy boundaries for enterprise deployment.
Classical ML
Solid grounding in scikit-learn and standard ML practice — feature engineering, model selection, cross-validation strategy, calibration and metric selection beyond accuracy.
Deep learning
Working knowledge of PyTorch or TensorFlow, and familiarity with fine-tuning approaches (LoRA / PEFT) and when they are and are not justified over prompting or retrieval.
Cloud
Production experience on Azure AI / Azure ML including working within governed single-tenant client environments.
Data fluency
Comfortable with SQL and distributed data processing; able to work directly against a lakehouse rather than waiting for someone to hand over a CSV.
Communication
Able to explain system behaviour, limitations and risk to non-technical stakeholders, and to make a model's reasoning credible to sceptical domain experts.
3.4 Good to Have
Model Context Protocol (MCP) or comparable standardised tool-integration approaches for connecting agents to enterprise systems.
Document intelligence — layout-aware parsing, OCR, table extraction, and multimodal document understanding at scale.
Fine-tuning, distillation or serving of open-weight models, including quantisation and inference-cost optimisation.
Time-series and signal-processing depth — filtering, resampling, spectral features — for sensor, telemetry or IoT engagements.
Geospatial machine learning — GPS traces, map matching, spatial indexing and segment-level aggregation.
Active learning, weak supervision or synthetic-data generation for label-scarce problems.
Knowledge graphs or ontology-driven data modelling as a complement to vector retrieval.
Front-end familiarity (React or similar) sufficient to prototype an agent or review interface without waiting on another engineer.
MLOps / LLMOps
Open-source contributions, published work, or demonstrable side projects in the agentic / LLM space.
3.5 Qualifications
Bachelor's or Master's degree in Computer Science, Data Science, Electrical / Electronics, Mathematics, Statistics or a related engineering discipline.
Relevant certification an advantage — Azure AI Engineer (AI-102), Azure Data Scientist (DP-100), AWS Machine Learning Specialty, GCP Professional ML Engineer or Databricks ML Professional.
Responsibilities
AI Engineer
Role Summary:
You design and ship the intelligence layer of our engagements. The centre of gravity is agentic AI — systems that plan across multiple steps, call tools and enterprise APIs, retrieve and reason over the client's own data, and complete real work rather than answering a single question. Around that sit the disciplines that make such systems trustworthy: retrieval design, evaluation harnesses, guardrails, observability and human-in-the-loop review.
Applied machine learning remains part of the role. Several engagements need forecasting, classification, anomaly detection or sensor-signal modelling rather than an agent, and the same person is expected to pick the right tool. A candidate who reaches for a language model regardless of the problem is not the profile we are hiring.
This is an engineering role, not a research role. The measure is systems that run in a client's production environment, survive contact with real users and real data, and can be explained to a sceptical business stakeholder — not benchmark scores or prototypes that never left a notebook.
3.2 Key Responsibilities:
Agentic System Design & Engineering
Design and build agentic systems that decompose goals into steps, select and call tools, use enterprise APIs and data sources, and recover sensibly when a step fails.
Define tool and function interfaces an agent can use reliably — clear contracts, validated structured inputs and outputs, sensible error surfaces, and idempotency where actions have side effects.
Implement orchestration and control flow: routing, planning, retries, escalation paths, state and memory management, and deterministic checkpoints where the business cannot tolerate variance.
Design human-in-the-loop and approval boundaries — deciding what an agent may do autonomously, what requires confirmation, and how a human takes over cleanly.
Build multi-agent workflows where they genuinely help, and resist them where a single well-scoped agent or a plain pipeline is the better answer.
Integrate agents with enterprise systems and identity — authentication, authorisation, least-privilege access to data and actions, and full auditability of what the agent did and why.
Retrieval & Knowledge Systems
Build retrieval-augmented systems over enterprise content — documents, tickets, contracts, manuals, wikis and structured data.
Own the retrieval quality chain: parsing and chunking strategy, metadata design, embedding selection, hybrid semantic and keyword search, reranking, and context assembly within practical limits.
Ground outputs in retrieved evidence with citations, and engineer explicitly against hallucination and confident-but-wrong answers.
Work with the Data Engineer on ingestion, incremental refresh, deletion propagation and index freshness so the knowledge layer does not silently drift out of date.
Evaluation, Guardrails & Reliability
Build evaluation harnesses before scaling a system, not after it fails: curated test sets, task-level success criteria, automated regression suites and model-graded evaluation where appropriate.
Define what 'good' means with the client in measurable terms, and report honestly against it — including where the system is weak and where it should not be trusted.
Implement guardrails: input and output validation, prompt-injection and jailbreak defence, PII detection and redaction, content filtering, and refusal behaviour for out-of-scope requests.
Instrument for observability — tracing multi-step runs, capturing tool calls and failures, tracking latency, token consumption and unit cost per task.
Manage cost and latency deliberately: model selection per task, caching, batching, context management, and cheaper fallbacks where full capability is not required.
Applied Machine Learning
Build classical and deep-learning models where they fit the problem better than a language model — forecasting, classification, regression, anomaly detection, clustering, time-series and sensor-signal modelling.
Engineer features from raw operational, transactional or sensor data, and design evaluation that reflects deployment reality — time-aware and grouped splits rather than random ones, and leakage avoidance.
Establish a credible baseline first, then justify added complexity with measured gain.
Work with domain experts to define labels and acceptance criteria in engagements where no ground-truth dataset exists at the outset.
Deployment, LLMOps & Handover
Deploy models and AI services into client environments — containerised services, batch scoring pipelines, or managed cloud AI platforms — with versioning, promotion criteria and rollback.
Monitor systems in production: data and prediction drift, quality regression, cost anomalies and failure patterns; own the retraining or re-tuning cycle.
Close the feedback loop — capture user and expert corrections and route them back into evaluation sets, retrieval improvements, prompt revisions or training data.
Make behaviour explainable to non-technical stakeholders, and deliver model documentation, evaluation reports and operational runbooks at handover.
Client-Facing Solutioning
Shape solutions with clients — assess AI feasibility for a business problem, size effort, identify data gaps, and say clearly when AI is the wrong tool for a given ask.
Demonstrate working increments on the client's own data in fortnightly sprint reviews, and support pre-sales technical discussions and proof-of-value work as required.
3.3 Must-Have Skills & Experience
Area
What We Expect
Experience
5–8 years in machine learning / AI engineering, including meaningful recent experience building LLM-based or agentic systems that reached production. Not research-only or notebook-only.
Core languages
Strong Python — production-quality, version-controlled, tested code. Comfortable building services and APIs (FastAPI or equivalent), not only models.
Agentic engineering
Demonstrable hands-on work with tool / function calling, multi-step agent loops, structured outputs, state and memory management, and failure recovery. Framework experience such as LangGraph, LangChain, LlamaIndex, CrewAI, AutoGen, Semantic Kernel or provider-native agent SDKs — with judgement about when a framework helps and when it gets in the way.
LLM application engineering
Practical depth with frontier model APIs (Anthropic, OpenAI, Google, or open-weight models via vLLM / Ollama / Bedrock): prompt design, structured output enforcement, streaming, context management, caching and cost / latency trade-offs.
Retrieval systems
End-to-end RAG delivery — chunking and metadata strategy, embedding models, vector databases, hybrid search, reranking, grounding and citation. Including knowing why a naive RAG implementation underperforms and how to fix it.
Evaluation
Building evaluation harnesses and regression suites for non-deterministic systems; defining task-level success metrics; model-graded evaluation and its limits; honest error analysis.
Guardrails & safety
Prompt-injection and jailbreak mitigation, input / output validation, PII handling, and designing appropriate autonomy boundaries for enterprise deployment.
Classical ML
Solid grounding in scikit-learn and standard ML practice — feature engineering, model selection, cross-validation strategy, calibration and metric selection beyond accuracy.
Deep learning
Working knowledge of PyTorch or TensorFlow, and familiarity with fine-tuning approaches (LoRA / PEFT) and when they are and are not justified over prompting or retrieval.
Cloud
Production experience on Azure AI / Azure ML including working within governed single-tenant client environments.
Data fluency
Comfortable with SQL and distributed data processing; able to work directly against a lakehouse rather than waiting for someone to hand over a CSV.
Communication
Able to explain system behaviour, limitations and risk to non-technical stakeholders, and to make a model's reasoning credible to sceptical domain experts.
3.4 Good to Have
Model Context Protocol (MCP) or comparable standardised tool-integration approaches for connecting agents to enterprise systems.
Document intelligence — layout-aware parsing, OCR, table extraction, and multimodal document understanding at scale.
Fine-tuning, distillation or serving of open-weight models, including quantisation and inference-cost optimisation.
Time-series and signal-processing depth — filtering, resampling, spectral features — for sensor, telemetry or IoT engagements.
Geospatial machine learning — GPS traces, map matching, spatial indexing and segment-level aggregation.
Active learning, weak supervision or synthetic-data generation for label-scarce problems.
Knowledge graphs or ontology-driven data modelling as a complement to vector retrieval.
Front-end familiarity (React or similar) sufficient to prototype an agent or review interface without waiting on another engineer.
MLOps / LLMOps
Open-source contributions, published work, or demonstrable side projects in the agentic / LLM space.
3.5 Qualifications
Bachelor's or Master's degree in Computer Science, Data Science, Electrical / Electronics, Mathematics, Statistics or a related engineering discipline.
Relevant certification an advantage — Azure AI Engineer (AI-102), Azure Data Scientist (DP-100), AWS Machine Learning Specialty, GCP Professional ML Engineer or Databricks ML Professional.
Salary : As per industry standard.
Industry :IT-Software / Software Services
Functional Area : IT Software - Application Programming , Maintenance
The Business Analyst will work closely with stakeholders, including product managers, developers, and healthcare professionals to analyze business needs and translate them into functional requirements. The ideal candidate will have 3-6 years of experience in the healthcare domain, proficiency in SQL, and a strong understanding of healthcare processes and regulations.
Key Responsibilities:
Gather, analyze, and document business requirements for healthcare projects.
Collaborate with cross-functional teams to design and implement healthcare solutions.
Perform data analysis using SQL to support decision making and reporting.
Understand healthcare workflows, compliance standards (e.g., HIPAA), and regulatory requirements.
Translate complex healthcare data and business needs into clear functional specifications.
Facilitate communication between technical teams and healthcare stakeholders.
Assist in user acceptance testing (UAT), validate deliverables against requirements.
Identify areas for process improvement and provide actionable recommendations.
Qualifications & Skills:
Bachelor’s degree in Business, Healthcare Informatics, Computer Science, or related field.
3 to 6 years of experience as a Business Analyst in the healthcare domain.
Strong proficiency in writing complex SQL queries for data extraction and analysis.
Good understanding of healthcare industry standards, EHR/EMR systems preferred.
Excellent analytical, problem-solving, and communication skills.
Experience with requirement gathering tools and techniques (e.g., JIRA, Confluence).
Ability to work in the shift timing of 2:00 PM to 11:00 PM IST.
Detail-oriented with the ability to manage multiple priorities effectively.
Preferred:
Experience with data visualization tools (e.g., Power BI, Tableau).
Knowledge of healthcare regulations such as HIPAA, ICD-10, CPT codes.
Exposure to Agile/Scrum methodologies.
Responsibilities
The Business Analyst will work closely with stakeholders, including product managers, developers, and healthcare professionals to analyze business needs and translate them into functional requirements. The ideal candidate will have 3-6 years of experience in the healthcare domain, proficiency in SQL, and a strong understanding of healthcare processes and regulations.
Key Responsibilities:
Gather, analyze, and document business requirements for healthcare projects.
Collaborate with cross-functional teams to design and implement healthcare solutions.
Perform data analysis using SQL to support decision making and reporting.
Understand healthcare workflows, compliance standards (e.g., HIPAA), and regulatory requirements.
Translate complex healthcare data and business needs into clear functional specifications.
Facilitate communication between technical teams and healthcare stakeholders.
Assist in user acceptance testing (UAT), validate deliverables against requirements.
Identify areas for process improvement and provide actionable recommendations.
Qualifications & Skills:
Bachelor’s degree in Business, Healthcare Informatics, Computer Science, or related field.
3 to 6 years of experience as a Business Analyst in the healthcare domain.
Strong proficiency in writing complex SQL queries for data extraction and analysis.
Good understanding of healthcare industry standards, EHR/EMR systems preferred.
Excellent analytical, problem-solving, and communication skills.
Experience with requirement gathering tools and techniques (e.g., JIRA, Confluence).
Ability to work in the shift timing of 2:00 PM to 11:00 PM IST.
Detail-oriented with the ability to manage multiple priorities effectively.
Preferred:
Experience with data visualization tools (e.g., Power BI, Tableau).
Knowledge of healthcare regulations such as HIPAA, ICD-10, CPT codes.
Exposure to Agile/Scrum methodologies.
Salary : As per industry standard.
Industry :IT-Software / Software Services
Functional Area : IT Software - Application Programming , Maintenance
As a Custom Software Engineer, a typical day involves designing and developing tailored software components that integrate seamlessly within larger systems or applications. The role requires continuous collaboration with team members to refine and enhance software solutions, ensuring they meet evolving business requirements. Utilizing contemporary development methodologies and frameworks, the engineer focuses on delivering efficient, scalable, and robust software that supports organizational goals. The position demands adaptability to changing project needs and active engagement in iterative development cycles to maintain high-quality outputs. Roles & Responsibilities: - Expected to perform independently and become an SME.- Required active participation/contribution in team discussions.- Contribute in providing solutions to work related problems.- Collaborate with cross-functional teams to gather and analyze requirements for custom software development.- Maintain and improve existing software components to enhance performance and reliability.- Document software designs, processes, and updates to ensure knowledge sharing and maintainability.- Assist junior team members by providing guidance and support in technical challenges. Professional & Technical Skills: - Must To Have Skills: Proficiency in SAP Quality Management (QM).- Strong understanding of software development life cycle and agile methodologies.- Experience in designing and implementing custom software solutions using modern frameworks.- Ability to troubleshoot and resolve complex software issues effectively.- Familiarity with integration of software components within enterprise systems.- Good communication skills to collaborate effectively with diverse teams. Additional Information: - The candidate should have minimum 3 years of experience in SAP Quality Management (QM).- This position is based at our Bengaluru office.- A 15 years full time education is required.
Responsibilities
As a Custom Software Engineer, a typical day involves designing and developing tailored software components that integrate seamlessly within larger systems or applications. The role requires continuous collaboration with team members to refine and enhance software solutions, ensuring they meet evolving business requirements. Utilizing contemporary development methodologies and frameworks, the engineer focuses on delivering efficient, scalable, and robust software that supports organizational goals. The position demands adaptability to changing project needs and active engagement in iterative development cycles to maintain high-quality outputs. Roles & Responsibilities: - Expected to perform independently and become an SME.- Required active participation/contribution in team discussions.- Contribute in providing solutions to work related problems.- Collaborate with cross-functional teams to gather and analyze requirements for custom software development.- Maintain and improve existing software components to enhance performance and reliability.- Document software designs, processes, and updates to ensure knowledge sharing and maintainability.- Assist junior team members by providing guidance and support in technical challenges. Professional & Technical Skills: - Must To Have Skills: Proficiency in SAP Quality Management (QM).- Strong understanding of software development life cycle and agile methodologies.- Experience in designing and implementing custom software solutions using modern frameworks.- Ability to troubleshoot and resolve complex software issues effectively.- Familiarity with integration of software components within enterprise systems.- Good communication skills to collaborate effectively with diverse teams. Additional Information: - The candidate should have minimum 3 years of experience in SAP Quality Management (QM).- This position is based at our Bengaluru office.- A 15 years full time education is required.
Salary : Rs. 0.0 - Rs. 1,60,000.0
Industry :IT-Software / Software Services
Functional Area : IT Software - Application Programming , Maintenance
Business Analyst - (260006N8)
Missions
Key Responsibilities
Collaborate effectively with team members and Paris counterparts while demonstrating the ability to work independently.
Own the delivery of key Business Analysis artifacts, including Functional Specifications, Wireframes, Root Cause Analysis (RCA), Source-to-Target Mapping, Test Strategy documents, and other project-required documentation.
Gather and analyze business requirements through discussions with business stakeholders to ensure clarity and alignment.
Translate user stories into detailed mapping documents for development and testing purposes.
Adhere to established project documentation standards and best practices.
Utilize strong hands-on SQL skills for data analysis and validation.
Analyze production data to derive meaningful KPIs and provide actionable insights for business users.
Proficiently use Jira for project tracking, task management, and collaboration.
Profile
Over 6 + years of experience in data-centric projects, including Data Warehouse and Data Lake implementations, preferably within the Banking domain.
Strong expertise in performing Gap Analysis and Root Cause Analysis to identify and resolve issues effectively.
Hands-on Business Analysis experience with proficiency in drafting detailed Functional Specifications.
Ability to translate business use cases into comprehensive Source-to-Target mapping sheets and perform functional validation.
Skilled in debugging production failures and providing accurate root cause solutions.
Solid understanding of SQL and RDBMS concepts with practical, hands-on experience.
Excellent analytical and troubleshooting skills to meet complex business requirements.
Familiarity with Agile methodologies and processes is an added advantage.
Effective team player with the ability to work autonomously and collaborate in cross-cultural environments.
Strong verbal and written communication skills for seamless interaction with stakeholders.
Responsibilities
Salary : As per industry standard.
Industry :IT-Software / Software Services
Functional Area : IT Software - Application Programming , Maintenance
