One non-cynical take on why modern software is slow, and not containing optimizations such as these: The standardization/optimization hypothesis.
If something is/has become a standard, then optimization takes over. You want to be fastest and meet all of the standard's tests. Doom is similarly now a standard game to port to any new CPU, toaster, whatever. Similarly email protocol, or a browser standard (WebRTC, Quic, etc).
The reason your latest web app/ electron app is not fast is that it is exploratory. It's updated everyday to meet new user needs, and fast-enough-to-not-get-in-the-way is all that's needed performance wise. Hence we see very fast IRC apps, but slack and teams will always be slow.
Kalvium is the B.Tech CS degree you always wanted. We redo 4000 hours of education, heavily utilizing GenAI to power self-directed education of all courses in undergraduate CS programs. This is the HEROS platform. We've grown from a team of ~30 in 2023 to ~300 today.
We're now undertaking HEROS v2, where there's a couple of very exciting challenges: How can GenAI be a 1-1 tutor during an hour of education. Can we keep the cost of the hour under 1 INR. Agentic workflows, vector-db/RAGs, finetuning and more techniques will be used.Interested in this challenge?
We're a team of deep technophiles, who are passionate about the transformation of education happening. We're partnered with over 2 dozen universities in India, and expanding overseas this year, growing 4-5X, year on year. And we've done all of this in a bootstrapped, default alive way.
This role would be one-off senior roles, working closely with head of engineer, and the co-founders in exploring and extending the edges of GenAI in education. Please send me an email note via my first name @kalvium.com if this sounds like something you'd like to take on, with evidence of work you've done with GenAI.
--
Competitive Coding Intern
Are you a 3rd or 4th year student, with a fantastic competitive coding profile? Interested in interning with us building our our library of coding questions. Reach out. Depending on your performance in the internship, we can also consider a full time role as you graduate.
Imo, this is a good proof of concept to show interviewers why current process is broken, and to evolve and fix it to be fairer. But the positioning by the author is offputting at best.
They're perhaps being penny wise and pound foolish with their career. Employment in tech is broadly out of the window for them - entrepreneurship is the only option.
> Security is not a differentiation method. It's table stakes for any technical product.
Security as table stakes, sure. SSO, certainly not.
It's an additional cost, last I checked it was between 10$ and 20$ per user per month if you take the cheapest option and outsource it.
This whole notion of "Unless you meet this security standard that 99% of products don't meet, your product hasn't met table stakes" is nonsense and needs to die.
SSO will get cheaper in the future; for now it's hard for a product development team to justify getting 0$ in revenue just because of some purity test by irrelevant folk on the internet.
Running my own keycloak or another ory hydra is a boring task. Locking SSO behind some arbitrary scale and raked up price takes sales away from you. It’s a matter of perspective.
> Running my own keycloak or another ory hydra is a boring task.
Simply running keycloak is not sufficient for SSO.
An SSO implementation may take months of dev time (i.e. $50k, minimum, considering cost of dev hours spent on it, and opportunity cost of not having those devs putting those hours into features).
And after you have done that, it remains an expensive feature - it's a high-touch feature that will eat product support time like you wouldn't believe.
Outsourcing this is still the cheapest option, and it still costs more than the product itself in many cases.
I don’t understand your point. My question is: if the service provider offers SSO as an additional feature, why limit it to certain size of a client? Their service supports it. Why cannot my two persons company enable this feature? My two persons company can run my own keycloak and use it as an OAuth provider in your product all right. If you need months of dev time to enable SSO on my account then say that upfront because I will certainly find a different service provider because you clearly don’t know what you’re doing.
> If you need months of dev time to enable SSO on my account then say that upfront because I will certainly find a different service provider because you clearly don’t know what you’re doing.
If you could do that, you would. The point is that SSO is high-touch and high-maintenance, and the price reflects that.
If it is as cheap you appear to think so, you'll make a killing offering SSO for businesses and undercutting the current providers by (maybe) 50%.
I don't think you are doing that. Maybe I am wrong, but if you are right you're leaving easy money on the table.
⍃Kalvium, Bangalore, India (onsite/hybrid - HSR Layout). Internships (DSA and coding Interns).
Hi folks, I'm hiring in the engineering curriculum team at Kalvium - the only edtech where we bring a focused, standardized, digital approach to CS higher education. I speak about this on this podcast (https://www.youtube.com/watch?v=be1yEJbyEkQ).
We have 6-month internships for DSA Programmers, and Project programmers. This role requires creating code/problems/projects for our students. Domains include full-stack, data-science, cybersecurity and DevOps. Post internship full-time conversion is also possible basis performance.
Here's 3 cool things we've done:
1. 100% AI based evaluators in B.Tech CS 4 year program.
2. Measuring and improving each one of the 4000 hours of the program in the classroom.
2. Launched our full-fledged program in 20+ universities in India, where students are placed from 3rd semester onwards. This year we've had more premium universities signup: SRM, VIT, Christ, etc.
3. Solid business fundamentals - sustainability in every classroom with our revenue.
Our main realization is: we can't rely on people to drive a significant change, rather it has to be a system which builds and uplifts skills in students. That's what we've built and launched in above universities (https://kalvium.com/heros/). We have a 12 year roadmap! it's a blue ocean, and we want to build and lead Edtech V2.0.
Please reach out to me via anil@kalvium.com with your resume. include your existing competitive coding platform links, coding projects, and any experience of tutoring others on technology. We have a high bar on this selection.
⍃Kalvium, Bangalore, India (onsite in HSR Layout). Internships, SWE2, Sr.SWE.
Hi folks, I'm hiring in the engineering team at Kalvium across SDE2 to Lead levels. We are building the google-maps of education - the only edtech where we bring a focused, standardized, digital approach to CS higher education. I speak about this on this podcast (https://www.youtube.com/watch?v=be1yEJbyEkQ).
At the entry level, we only have final year internships (9-12 months) + PPO for students who are graduating in 2025. We index heavily on prior project/open source work for this role.
Here's 3 cool things we've done:
1. 100% AI based evaluators in B.Tech CS 4 year program.
2. Measuring and improving each one of the 4000 hours of the program in the classroom.
2. Launched our full-fledged program in 20+ universities in India, where students are placed from 3rd semester onwards. This year we've had more premium universities signup: SRM, VIT, Christ, etc.
3. Solid business fundamentals - sustainability in every classroom with our revenue.
To align you to our approach: we seek out engineers who want to involve with this problem statements. Edtech v1.0 in India does not have a good rep (tell me about it!), but the problem statements remains: higher education system is broken, and not giving students the outcomes they are capable of (Here's the outcomes we deliver - https://kalvium.com/hire-from-us/).
Our main realization is: we can't rely on people to drive a significant change, rather it has to be a system which builds and uplifts skills in students. That's what we've built and launched in above universities (https://kalvium.com/heros/). We have a 12 year roadmap! it's a blue ocean, and we want to build and lead Edtech V2.0.
Note on team culture/equity: We plan on being a small core team that geeks out on learning and platform-izing it! Our pay is commensurate to experience with meaningful equity. We plan for a 4 year path of growth for every individual in our team. If this sounds like a meaningful challenge and place to build your career, please reach out to me via anil@kalvium.com with your resume.
I don't know if you meant it (apologies if not) - but libraries, and CC, and FOSS software, and Wikipedia are all good things in humanity. We should aim for more such good things.
"Salted" or not -- it's up to supporters of free markets/capitalism to figure their shit out.
I very much agree that the things you mention are all good things. Condemning them in general is not what I meant, though I can see how my comments could be read that way.
No, my point is more specific: it's that those things play with the free market about as well as NaN plays with floating point math. That by itself isn't bad; the market isn't the best answer to everything. However, in case of F/OSS, I wish people acknowledged that, by destroying the ability to just sell software on a free market (including software components), it's in big part responsible for today's SaaS-ified software reality.
> ... by destroying the ability to just sell software on a free market (including software components), it's in big part responsible for today's SaaS-ified software reality.
Could you please elaborate on this or point me to a source where that exact mechanism is explained? Because this runs somewhat opposite to my experiences where FOSS was more of a desperate way to escape proprietary software, particularly OS like IBM with which you would have to wait for fixes from "the market" for days instead of being able to fix it yourself, like this article explains (in vastly superior English to mine): https://cacm.acm.org/practice/free-and-open-source-software-...
In education at least, we've actively improved efficiency by ~25% across a large swath of educators (direct time saved) - agentic evaluators, tutors and doubt clarifiers. The wins in this industry are clear. And this is that much more time to spend with students.
I also know from 1-1 conversation with my peers in large-finance world, and there too the efficiency improvements on multiple fronts are similar.
They are partially hype though. That's what people here are arguing. There are benefits but their valuation is largely hype driven. AI is going to transform industries and humanity, yes. But AI does not mean LLM (even if LLM means AI). LLM raw potential was reached last year with GPT-4. From here on, the value will lie on exploiting the potential we already have to generate clever applications. Just like the internet provided a platform for new services, I expect LLMs to be the same but with a much smaller impact
If something is/has become a standard, then optimization takes over. You want to be fastest and meet all of the standard's tests. Doom is similarly now a standard game to port to any new CPU, toaster, whatever. Similarly email protocol, or a browser standard (WebRTC, Quic, etc).
The reason your latest web app/ electron app is not fast is that it is exploratory. It's updated everyday to meet new user needs, and fast-enough-to-not-get-in-the-way is all that's needed performance wise. Hence we see very fast IRC apps, but slack and teams will always be slow.