Connect with us

AI in Travel

A Gentle Introduction to Context Engineering in LLMs

Published

3 days ago

on


Context Engineering Overview VisualImage by Author | Canva

 

# Introduction

 
There is no doubt that large language models can do amazing things. But apart from their internal knowledge base, they heavily depend on the information (the context) you feed them. Context engineering is all about carefully designing that information so the model can succeed. This idea gained popularity when engineers realized that simply writing clever prompts is not enough for complex applications. If the model doesn’t know a fact that’s needed, it can’t guess it. So, we need to assemble every piece of relevant information so the model can truly understand the task at hand.

Part of the reason the term ‘context engineering’ gained attention was due to a widely shared tweet by Andrej Karpathy, who said:

+1 for ‘context engineering’ over ‘prompt engineering’. People associate prompts with short task descriptions you would give an LLM in your day-to-day use, whereas in every industrial-strength LLM app, context engineering is the delicate art and science of filling the context window with just the right information for the next step…

 

This article is going to be a bit theoretical, and I will try to keep things as simple and crisp as I can.

 

# What Is Context Engineering?

 
If I received a request that said, ‘Hey Kanwal, can you write an article about how LLMs work?’, that’s an instruction. I would write what I find suitable and would probably aim it at an audience with a medium level of expertise. Now, if my audience were beginners, they would hardly understand what’s happening. If they were experts, they might consider it too basic or out of context. I also need a set of instructions like audience expertise, article length, theoretical or practical focus, and writing style to write a piece that resonates with them.

Likewise, context engineering means giving the LLM everything from user preferences and example prompts to retrieved facts and tool outputs, so it fully understands the goal.

Here’s a visual that I created of the things that might go into the LLM’s context:

 

Context Engineering DiagramContext Engineering Diagram Context engineering includes instructions, user profile, history, tools, retrieved docs, and more | Image by Author
 
 

Each of these elements can be viewed as part of the context window of the model. Context engineering is the practice of deciding which of these to include, in what form, and in what order.

 

# How Is Context Engineering Different From Prompt Engineering?

 
I will not make this unnecessarily long. I hope you have grasped the idea so far. But for those who didn’t, let me put it briefly. Prompt engineering traditionally focuses on writing a single, self-contained prompt (the immediate question or instruction) to get a good answer. In contrast, context engineering is about the entire input environment around the LLM. If prompt engineering is ‘what do I ask the model?’, then context engineering is ‘what do I show the model, and how do I manage that content so it can do the task?’

 

# How Context Engineering Works

 
Context engineering works through a pipeline of three tightly connected components, each designed to help the model make better decisions by seeing the right information at the right time. Let’s take a look at the role of each of these:

 

// 1. Context Retrieval and Generation

In this step, all the relevant information is pulled in or generated to help the model understand the task better. This can include past messages, user instructions, external documents, API results, or even structured data. You might retrieve a company policy document for answering an HR query or generate a well-structured prompt using the CLEAR framework (Concise, Logical, Explicit, Adaptable, Reflective) for more effective reasoning. 

 

// 2. Context Processing

This is where all the raw information is optimized for the model. This step includes long-context techniques like position interpolation or memory-efficient attention (e.g., grouped-query attention and models like Mamba), which help models handle ultra-long inputs. It also includes self-refinement, where the model is prompted to reflect and improve its own output iteratively. Some recent frameworks even allow models to generate their own feedback, judge their performance, and evolve autonomously by teaching themselves with examples they create and filter.

 

// 3. Context Management

This component handles how information is stored, updated, and used across interactions. This is especially important in applications like customer support or agents that operate over time. Techniques like long-term memory modules, memory compression, rolling buffer caches, and modular retrieval systems make it possible to maintain context across multiple sessions without overwhelming the model. It is not just about what context you put in but also about how you keep it efficient, relevant, and up-to-date.

 

# Challenges and Mitigations in Context Engineering

 
Designing the perfect context isn’t just about adding more data, but about balance, structure, and constraints. Let’s look at some of the key challenges you might encounter and their potential solutions:

  • Irrelevant or Noisy Context (Context Distraction): Feeding the model too much irrelevant information can confuse it. Use priority-based context assembly, relevance scoring, and retrieval filters to pull only the most useful chunks.
  • Latency and Resource Costs: Long, complex contexts increase compute time and memory use. Truncate irrelevant history or offload computation to retrieval systems or lightweight modules.
  • Tool and Knowledge Integration (Context Clash): When merging tool outputs or external data, conflicts can occur. Add schema instructions or meta-tags (like @tool_output) to avoid format issues. For source clashes, try attribution or let the model express uncertainty.
  • Maintaining Coherence Over Multiple Turns: In multi-turn conversations, models may hallucinate or lose track of facts. Track key information and selectively reintroduce it when needed.

Two other important issues: context poisoning and context confusion have been well explained by Drew Breunig, and I encourage you to check that out.

 

# Wrapping Up

 

Context engineering is no longer an optional skill. It is the backbone of how we make language models not just respond, but understand. In many ways, it is invisible to the end user, but it defines how useful and intelligent the output feels. This was meant to be a gentle introduction to what it is and how it works.

If you are interested in exploring further, here are two solid resources to go deeper:

 
 

Kanwal Mehreen is a machine learning engineer and a technical writer with a profound passion for data science and the intersection of AI with medicine. She co-authored the ebook “Maximizing Productivity with ChatGPT”. As a Google Generation Scholar 2022 for APAC, she champions diversity and academic excellence. She’s also recognized as a Teradata Diversity in Tech Scholar, Mitacs Globalink Research Scholar, and Harvard WeCode Scholar. Kanwal is an ardent advocate for change, having founded FEMCodes to empower women in STEM fields.



Source link

Related Topics:
Continue Reading
Click to comment

You must be logged in to post a comment Login

Leave a Reply

AI in Travel

AI-Powered Travel: UAE Leads in Smart, Seamless Experiences

Published

15 hours ago

on

August 10, 2025

By


In a country that’s synonymous with futuristic skylines and hyper-connected infrastructure, it’s no surprise that the UAE is redefining what it means to travel. From the moment a trip is planned to the final post-travel review, artificial intelligence (AI), biometrics, and automation are increasingly at the heart of the experience. Airports are becoming smarter, travel planning is more personalised than ever, and reliability is being re-engineered by the minute.

A New Era of Intelligent Travel

Dubai International Airport (DXB) is leading the charge when it comes to next-gen travel experiences powered by AI. As the world’s busiest international hub, the pressure to deliver smooth, secure, and swift passenger journeys is immense and technology is rising to the occasion. “AI is revolutionising the UAE travel journey from start to finish,” says Omar Bin Adai, Chief Technology and Infrastructure Officer of Dubai Airports. “At DXB, biometric smart gates offer seamless passport control with facial verification, eliminating manual checks. Our new ‘Unlimited Smart Travel’ takes this further, enabling up to 10 guests to complete immigration in just 14 seconds using facial recognition alone.”

Beyond passenger flow, AI is working hard behind the scenes. According to Adai, predictive maintenance and AI-driven baggage systems are ensuring near-perfect operations. “In Q1 2025 alone, DXB processed over 21 million bags with a 99.8% accuracy rate and one of the world’s lowest mishandling rates – 1.95 per 1,000 passengers.”

These numbers aren’t just impressive—they represent a growing commitment to using data and intelligence to deliver exceptional guest experiences. “This strategic integration of AI across every touchpoint exemplifies how the UAE is setting a global standard for intelligent, customer-centric travel infrastructure,” Adai adds.

While airports are getting smarter, so too is the way travellers plan their trips. Platforms powered by generative AI such as ChatGPT and Gemini are transforming the discovery phase of travel, giving users the ability to explore destinations, craft itineraries, and make informed decisions faster than ever before.

“We’re seeing strong uptake of AI-powered tools that support travellers, particularly during the research and planning phases of a trip,” says John Bevan, CEO of dnata Travel Group. “Platforms like ChatGPT have made travel information more accessible, helping users generate ideas and structure itineraries in seconds.”

Still, the technology has room to grow. “These platforms can struggle with the complex logistics of a full travel journey – managing bookings across multiple suppliers, handling real-time availability, and dynamic pricing. But the potential is enormous, especially in hyper-personalisation,” Bevan notes.

At dnata, the focus is on using AI to augment human expertise, not replace it. “We’re excited about this future. We’re integrating AI not just for efficiency, but to amplify the knowledge of our travel consultants – making it faster and smarter to deliver relevant, high-quality advice at scale.”

Elevating the Travel Experience

Technology is reshaping customer service like never before. Mobile apps, chatbots, and virtual assistants have become vital touchpoints, offering travellers personalised, real-time support that transforms uncertainty into confidence. Bin Adai highlights the profound impact: “In today’s travel landscape, the real game-changer has been placing instant, personalised support directly into passengers’ hands. Mobile apps and virtual assistants, particularly intuitive tools like DXB’s wayfinding app, have reshaped how travellers interact with airports.”

Among these innovations is DXB Express Maps, a dedicated app for Dubai International Airport that provides travellers with interactive 3D maps and a user-friendly interface to easily find gates, dining options, and other facilities. “Instead of uncertainty, travellers now effortlessly navigate complex terminals, access timely flight information, and manage their journeys seamlessly,” Bin Adai adds.

Complementing this is Pocket Flights, an app that gives instant access to real-time flight updates by simply scanning a QR code on flight status screens. Available in both English and Arabic across DXB and DWC airports, it delivers comprehensive information including gate changes, walking distances, wait times, and departure details — all at travellers’ fingertips.

The Airport Community App, affectionately dubbed the “mini-AOCC in your pocket,” supports over 59,000 users across 170+ entities in the oneDXB community. It merges user-driven features with operational tools designed to enhance guest experience and streamline airport efficiency. “This initiative strengthens internal communication and positions the app as an essential daily platform for the entire airport community,” Bin Adai explains.

Looking ahead, AI-driven recommendation engines are set to raise the bar even higher. “These systems are increasingly sophisticated, anticipating traveller preferences with precision and offering tailored, trusted suggestions,” says Bin Adai. “Ultimately, it’s about empowering the traveller with technology that feels human and intuitive, ensuring they always feel supported, understood, and confident at every step.”

Bevan echoes this sentiment but underscores the balance between speed and reliability: “Customer service today is increasingly being defined by immediacy and availability, especially in the travel sector where support is expected instantaneously – whether a person is booking a flight at midnight or making last-minute changes during a layover.”

While dnata is exploring consumer-facing chatbots as part of its future roadmap, its current AI investments focus on backend operations. “We’re already using AI to support training and quality assurance, including automated call listening capabilities that help us monitor service levels, identify pain points, and continuously improve our customer experience,” Bevan explains.

He adds that AI is also automating time-consuming manual processes, freeing teams to engage in higher-value interactions. “As we continue evaluating advanced communications technologies, accuracy and reliability remain key considerations. It’s not just about speed – it’s about ensuring the tools can provide relevant, context-aware support.”

Smart Journeys Ahead

No longer content with just convenience and punctuality, today’s traveller expects more — more control, more personalisation, and a more immersive experience from the moment a trip is imagined to the moment it ends.

According to Bin Adai, the shift is unmistakable. “Travellers in the UAE are increasingly expecting journeys that are not just seamless, but also smart, personalised, and immersive,” he says. “There’s a clear shift from traditional travel touchpoints to tech-enabled experiences that feel intuitive and engaging.” This evolution is powered by technologies like facial recognition for smoother immigration, AI-generated travel suggestions, and mobile-first tools that allow real-time support and bookings at the swipe of a screen. Airports like DXB have already embraced wayfinding apps and digital integration to deliver convenience with minimal friction.

But the transformation doesn’t end at the airport gates. “Travellers want digital convenience from planning to post-travel feedback,” Bin Adai notes. “Mobile-first platforms, immersive booking tools, and real-time support are no longer luxuries; they’re the baseline.”

This growing appetite for tech-forward travel solutions is not without its caveats. While AI and automation are streamlining the experience, the demand for human-centric service remains strong. “Technology is elevating convenience and efficiency, but the human touch remains essential,” he adds. “Dubai Airports continues to prioritise hospitality and in-person support, ensuring that travellers receive empathetic, culturally attuned service alongside advanced digital solutions.”

While echoing the sentiment, Bevan highlights an important distinction: modern travellers aren’t only looking for fast and easy experiences, they’re looking for assurance and options. “Travellers in the UAE today expect more than just convenience – they want control, flexibility, and confidence throughout their journey,” he explains. “What we’re seeing is a shift toward multi-channel and tech-enabled planning, where people want the freedom to engage with travel brands on their own terms — whether that’s online, in-store, through an app, or over the phone.”

For providers like dnata, that means maintaining consistency across all platforms and understanding that one solution won’t fit all. “Different age groups and lifestyles require different touchpoints, and it’s our responsibility to meet those needs with consistency, safety, and reliability,” says Bevan. “That means offering a seamless, secure experience whether a customer is booking through a chatbot at midnight or sitting down with an agent in one of our retail shops.”

He also underscores a key concern in this new era of digital interaction: trust. “As travellers share more information to receive personalised recommendations, they also expect us to handle that data with the highest levels of security,” Bevan notes. “Ultimately, it’s about building confidence, combining immersive and tech-integrated tools with the trust and assurance people need to feel good about their travel choices.”

What’s Next?

Over the next five to ten years, the country is poised to lead a radical shift toward intelligent, hyper-connected travel experiences that prioritise both ease and personalisation. “Over the next decade, the UAE is poised to lead a global shift toward smarter, fully connected travel,” says Bin Adai. “Biometric and contactless journeys will soon become the norm, enabling passengers to move through airports without ever presenting a document.” This vision is being woven into the very blueprint of the UAE’s next aviation mega-project — the expansion of Dubai World Central – Al Maktoum International (DWC). “The new airport will set a new benchmark for efficiency, capacity, and traveller-centric design,” Bin Adai adds.

From personalised service delivery and proactive customer support to predictive maintenance and operational efficiency, AI will touch every corner of the travel experience. “Our new digital experience project will cater to the latest AI technologies, helping elevate guest services to an entirely new level,” says Bin Adai. Looking further ahead, he sees the seamless integration of autonomous air taxis, high-speed ground transit, and eco-conscious infrastructure as part of a broader mobility ecosystem cementing the UAE’s place as a global hub for future-ready travel.

Bevan shares a similarly bold outlook. “Looking ahead, we see technologies like AI-generated itineraries, biometric-enabled travel, and predictive pricing engines becoming standard across the UAE’s travel landscape,” he says. “These tools are already beginning to take shape and in a region as digitally advanced and globally connected as the UAE, adoption will only accelerate.”



Source link

Continue Reading

AI in Travel

MakeMyTrip launches AI-enabled travel agent Myra

Published

1 day ago

on

August 9, 2025

By


Online travel booking platform MakeMyTrip has introduced an AI-powered virtual travel agent that can guide users through every step of their journey on the website, from trip planning and booking to handling post-sales queries such as cancellations and refunds, via both voice and text.

Users can ask complex and open-ended queries in the realm of travel in Hindi or English like “Where can I go in August for a relaxing holiday with my kids?,” or “I want to go to south India to cover Madurai, Rameswaram, Kovalam, Kodaikanal. Can you suggest me the best route? But I don’t want to travel via flight”.

The virtual assistant Myra is built on a network of specialised AI agents across all major travel categories, flights, accommodation, holidays, ground transport, visas, and forex. It supports multimodal input (text, voice, image, video), continuous back-and-forth dialogue, itinerary edits, and post-sales support — all within the same interface.

Myra will gives user personalised answers based on up-to-date availability, prices, and relevance. MakeMyTrip claims that while most AI travel tools only offer suggestions to users, this tool goes a step further by helping users move from travel ideas to actually helping travellers book their travel and complete payment online – something that hasn’t been tried before. It will offer assistance not just with flights and hotels but also holiday planning, ground transport such as cabs and buses, visas and forex.

CEO Rajesh Magow said such digital innovations helped “reach the deepest corners, and bringing seamless, intelligent travel booking to those who have long been underserved by digital platforms.”

“MakeMyTrip has seen travel demand penetration grow deeper over the years, from metro to top 15 cities many years ago to now as many as 2,000 unique cities,” he told The Hindu. The portal commands more than 50% market share among all online travel booking portals.

Myra may be able to facilitate bookings for an individual but it lacks the complexities of negotiating, handling customer bargains, accounting for individual preferences for a group bookings like a human travel agent, or helping during a travel emergency.

“Even if AI is able to do 50% of a human tour manager’s work, there will still be room for workforce in other tasks and AI related human intervention,” he added.

The tool is currently available in Hindi and English and will be expanded to include other regional languages too. It is currently available in beta version which will be released to a limited group of real users for testing before the official launch.

Group Chief Technology Officer Sanjay Mohan called the product “the most ambitious build” undertaken by the company.

In the next stage, MakeMyTrip will add smarter search tools that can understand the meaning behind what you’re looking for, even from images and videos. This means one won’t have to rely only on fixed filters and can search in a more intuitive manner. The system will be able to pick up on subtle travel needs, making it easier and more personal to find what you want.

Published – August 09, 2025 10:01 pm IST



Source link

Continue Reading

AI in Travel

Indian Railways Now to Launch AI Centre of Excellence for Smarter and Safer Rail Operations

Published

1 day ago

on

August 9, 2025

By



Published on
August 9, 2025 |

In a notable move toward the modernization of India’s railway system, Indian Railways has announced plans to set up an Artificial Intelligence (AI) Centre of Excellence (CoE). This new center will focus on researching and AI application development for the optimization of railway operations, enhanced safety for passengers, improved service efficiency, and the futurization of Indian Railways. The CoE will focus on using AIs impact on operational efficiency and will serve as a key institution to propel India’s vision of incorporating advanced technology in public transportation systems.

The impact of the Railway AI Centre of Excellence will form a core portion of Indian Railways’ ambitious modernization plans which features a holistic approach that merges systematic plans to develop the railway infrastructure with the modernization of services across the country’s vast railway network. Indian Railways plans to change the use of AI by introducing advanced technology in the management of core functions such as train operations and passenger services. The government sees this as a critical step in enhancing service quality.

Purpose of the AI Centre of Excellence

The primary focus of the Railway AI Centre of Excellence is to harness the capabilities of AI in automating essential functions of railway operations. Problems of Indian Railways that are multifaceted in nature and are solved on a daily basis will be addressed by the Centre using sophisticated machine learning techniques, predictive analytics, and AI-based decision systems.

The AI CoE also aims to enhance safety on the railway tracks. Indian railway is one of the largest railway systems in the world, and the safety of passengers and cargo is critically important. AI technologies will be instrumental in improving the accuracy of safety operations and can be integrated into the railway systems for real-time data analysis, predictive maintenance, and the automation of accident detection systems to negate human error, streamline operations and avoid accidents.

AI in Train Operations and Maintenance

The AI Centre will concentrate on optimizing train operations and maintenance. With AI-driven predictive maintenance tools, Indian Railways will be able to monitor the state of trains, tracks, and all other infrastructure in real time. This will drastically cut downtime due to mechanical failures, and assist railway operators in proactive issue mitigation.

AI predictive algorithms will also enhance train scheduling and route selection based on real-time updates. This will improve the punctuality of trains and aid in the alleviation of congestion at the stations, thus improving traffic management on the tracks.

AI will improve efficiency in the management of train yards, signaling systems, and crew schedules. This will enhance the operational and systemic efficiency of the AI tools will enable better tracking and analysis of train movements, enable pattern tracking to foresee peak travel times, and subsequently enable better scheduling decisions.

Enhancing Passenger Experience Through AI

The AI Centre of Excellence prioritizes enhancing passenger safety and comfort at every stage of travel. AI can elevate customization and personalization in passenger interactions and services. The ticketing process can be simplified and secured with face recognition and AI algorithms, offering smart ticketing.

The AI Centre will also deploy sophisticated AI solutions to manage passenger crowds at station and platform levels. Through AI surveillance and monitoring, foot traffic, and predicted crowd movement, station management will be able to manage passenger traffic with more precision, minimizing congestion.

To improve customer support services, AI-powered chatbots and virtual assistants will be able to attend to passenger queries in real-time. Other AI-driven systems will be tasked with functions related to comfort during travel, including real-time detection of seat availability and suggesting routes tailored to passenger preferences.

Boosting Operational Efficiency and Cost Reduction

As one of the biggest employers in the world, the Indian Railways has a holistic impact on the nation’s economic growth and productivity . Streamlining operations using AI systems will cut operational costs and eliminate inefficiencies in scheduling, fuel management, and employee positioning.

AI will ensure that goods and resources are available at the necessary railway stations, automating the food, water, and essential supply systems so that they can be monitored and replenished in real time. This will guarantee that passengers receive the best possible service.

Embracing AI will not only improve the operational efficiency of the railway network, but will pave the way for realized savings. These saved costs can be channeled back into the railway infrastructure for further investment.

Collaboration with Technology Partners

To accomplish this ambitious endeavour, Indian Railways will partner with the top technology companies and AI specialists. The government has mentioned that the AI Centre of Excellence will collaborate with universities, and other innovation hubs to establish an inter-disciplinary framework for synergy. This partnership will allow Indian Railways to take advantage of the breakthroughs in Artificial Intelligence and the newest technology in their operations.

The Indian government will fully finance and allocate resources for the creation of AI applications to be integrated throughout the Indian Railways network. The initiative will also establish an ecosystem for the creation of specialized AI applications and services intended for the Indian railway ecosystem.

Indian Railways’ Role in Digital Transformation

The foundation of the AI Centre of Excellence marks a milestone for Indian Railways in the domain of technology integration and the digital transformation journey. The Centre is expected to be a game changer in the use of AI in public transport and is in line with the government’s “Digital India” initiative which strives to bring a digital overhaul in different spheres of the economy.

This is in line with the vision of Narendra Modi to strive for India as a globalsuperpower in technology and innovation. The Indian Railways is looking to position themselves as a leader in the integration of technology with public transportation with AI at the forefront of the next wave of growth.

The Road Ahead

The Railway AI Center of Excellence is set to begin operations in the near future, with the intention to roll out AI technologies throughout the railway ecosystem over time. The government has underscored that the initiative’s success will hinge not just on technology, but also on the proper human resource who will be tasked with operating and managing the AI technology.

The Railway AI Center of Excellence will, in the long term, be instrumental in the Indian Railways’s transformation. Enhancing the network’s intelligence, safety, and operational efficiency is poised to transform railway travel for the millions of passengers who travel across the country on a daily basis.



Source link

Continue Reading

Trending

Copyright © 2025 AISTORIZ. For enquiries email at prompt@travelstoriz.com