Scaler Companion — Three students, one amazing Gen AI tool
What inspired you to develop Scaler Companion, and what specific problems were you aiming to solve for learners on the Scaler platform? What makes it innovative?
We were inspired to develop Scaler Companion after seeing the dedication of working professionals in Scaler’s upskilling programs when preparing for tech interviews. At Scaler School of Technology (SST), we’ve had the privilege of learning from instructors who’ve worked at Amazon, Google, and Microsoft. Their insights into top companies' interview processes showed a real need for accessible, focused preparation tools to make interview prep more efficient and personalised to bridge the gap between learning and readiness for tech roles. We wanted to help professionals who upskill through Scaler and our peers at SST.
So, we combined our skills: Ayaan on AI integration, Abhinav on backend development, and Pritam handling frontend design. To make it even more impactful, we incorporated the voice of Scaler’s Co-founder, Anshuman Singh, into the mock interview feature so that the tool could carry his knowledge and insights into every user’s upskilling journey.
Can you walk us through the Note Maker feature? How does it work to enhance the learning experience?
The Note Maker feature in Scaler Companion was designed to make learning more effective and accessible. After each live session, it gathers lecture transcripts, instructor notes, and related reading materials, consolidating them into concise, personalised flashcards for quick review. This helps Scaler learners quickly absorb the core concepts covered in class, which is crucial when preparing for interviews or completing assignments. We wanted something that allowed learners to revisit critical information without going through lengthy notes.
The Note Maker feature promotes better retention by breaking down complex materials into easy-to-review flashcards. It makes learning an ongoing process, helping students stay focused and prepared at each step of their journey.
What kind of customisation options does the note maker offer to adapt to different learners' needs? Can students highlight or prioritise certain lecture parts for more detailed summaries?
The Note Maker automatically identifies key points from lectures and generates concise, personalised notes to ensure all learners receive consistent, high-quality summaries. While users cannot directly customise these notes, the Companion includes an AI support bot to assist with clarifications and answer specific questions, providing a layer of personalised learning support.
Looking ahead, we’re exploring exciting customisation features to enhance the learning experience further. One potential enhancement is allowing students to highlight specific points or lines within the notes, enabling them to generate detailed, tailored summaries or initiate personalised discussions with the AI.
This feature would give learners more control over their study process, helping them focus on the most critical areas. While these capabilities are still in development, they represent our commitment to making the Note Maker a genuinely interactive and adaptive learning tool.
How did you design the mock interview feature to simulate real-world interviews, and why did you model the interviewer’s avatar after Scaler’s co-founder?
Our mock interview feature was designed to give students a realistic and interactive experience. We used advanced speech-to-speech interaction so that users could engage in spoken dialogue, creating a setting closer to a live interview. To further enhance this experience, we modelled the AI avatar after Scaler’s Co-founder, Anshuman Singh.
This choice was intentional — we wanted to create a tool that could embody Anshuman’s expertise and know-how to support learners on their upskilling journey best. We felt he would add a personal touch that resonates with the Scaler community, helping students feel more connected and engaged as they prepare for high-stakes interviews.
How was the mock interview module designed to mimic the interview styles of major tech companies like Google and Meta? Please share some details on the tools and scenarios the module was trained on.
The mock interview module was crafted to reflect the unique interview processes of major tech companies like Google and Meta by focusing on company-specific question styles, formats, and scenarios.
For instance, it includes questions on data structures and algorithms, behavioural queries, and system design to prepare students with a comprehensive and targeted interview practice experience. We also received valuable insights from Anshuman, who conducted over 500 hiring interviews with prestigious educational institutes at Meta’s London office.
We were able to tailor the module to simulate the types of questions, problem-solving approaches, and interaction styles typical of these companies. We used OpenAI and Claude models for the technical framework to support natural language interaction and GROQ speech-to-text and Play HT’s voice features to ensure responsive dialogue.
What technical or logistical challenges did you encounter while building the Scaler Companion, and how did you overcome them?
Building the Scaler Companion came with several challenges. On the technical front, ensuring accuracy by reducing AI hallucinations, minimising response latency, and designing a modular architecture for seamless platform integration were critical hurdles.
Logistically, calibrating the difficulty levels of mock interview simulations and addressing gaps to improve evaluation strictness posed significant challenges. Security concerns, such as enabling proctoring and preventing tab changes during assessments, were key considerations that required careful optimisation. These efforts have shaped Scaler Companion into a reliable and impactful tool for learners.
What feedback have you received from learners and instructors who have used Scaler Companion? Do you have any insights on its impact?
Feedback from both learners and instructors on the Scaler Companion has mainly been positive. Scaler learners appreciate AI-driven support, which provides quick, accurate answers to questions and reduces the need for human support by 30%. With automatic updates on Scaler-related content, the Companion stays relevant and practical, helping users stay engaged and informed.
Instructors have also noted a boost in efficiency, as real-time support from the Companion reduces their workload and lets them focus on more complex teaching tasks. This dual benefit highlights the tool's positive impact on both learning outcomes and instructional processes within Scaler.
Are there any other innovations or projects you’re currently working on, and how do they align with your long-term goals in tech and AI?
Pritam, Abhinav, Ayaan, and other SST students are working on several inspiring projects. However, it is still too early to discuss most of them. All these projects align with their shared vision of using technology to address societal challenges, empower individuals, and foster innovation on a broader scale.
How do you envision expanding Scaler Companion or evolving your work with generative AI tools to benefit the broader tech education space?
Looking ahead, we plan to expand Scaler Companion by building an AI avatar modelled on Scaler’s Co-founder, Anshuman Singh. Designed to embody his knowledge and experience, this avatar will look and sound like Anshuman sir, offering learners teaching assistance and interview preparation tailored to his expertise. This innovation aims to make tech education even more interactive and personalised, allowing learners to engage with a virtual mentor as they build their skills.
